YouTube2Text

The Quant Who Studied 700 Yrs Of Price Data: Most Levels Are Noise! THESE Are The 2 Levels To Watch! — Transcript

by Titans Of Tomorrow · 18,623 words · 2,802 segments · language en · Watch on YouTube

Full transcript

  1. 0:00This quant traded orders big enough to
  2. 0:02move the market, spent decades
  3. 0:04researching 700 years of price data and
  4. 0:07sold what he found to some of the
  5. 0:09biggest hedge funds on Wall Street. And
  6. 0:12today you get it here for free.
  7. 0:14>> One thing that makes my work almost
  8. 0:16unique is I've looked at the full scale
  9. 0:19of available data in financial markets
  10. 0:21all the way back to the Phoenicians. The
  11. 0:22way that I trade, that's a fundamental
  12. 0:24principle that has existed as long as
  13. 0:26we've had financial markets. On the
  14. 0:28other hand, if I were to decide to
  15. 0:30compute the Jane Street, now I have a
  16. 0:31problem because the length of a copper
  17. 0:34cable, the speed of light through that
  18. 0:35cable matters. They're co-located. I'm
  19. 0:37at home. I brought the wrong tool for
  20. 0:40the job and I can't compute on that time
  21. 0:42frame.
  22. 0:42>> Introducing Adam Grimes, former New York
  23. 0:44Mercantile Exchange and SMB Capital
  24. 0:47trader, author of The Art and Science of
  25. 0:50Technical Analysis, and he's written and
  26. 0:52published institutional research every
  27. 0:54day for nearly a decade.
  28. 0:56>> People have been looking at order flow
  29. 0:57ever since markets went electronic and
  30. 1:00surely there's information there, but
  31. 1:01there's also a tremendous amount of
  32. 1:03deception on the tape. I've been on both
  33. 1:04sides of that. I've been reading the
  34. 1:06tape and I've also been the guy trying
  35. 1:08to screw the people reading the tape. I
  36. 1:10know the tricks to do that.
  37. 1:11>> What does that mean, the second part?
  38. 1:12>> Well, it means that
  39. 1:15let's talk about Fibonacci specifically.
  40. 1:16Does the market reverse at that level
  41. 1:18first of all? That's the question. How
  42. 1:20would we test that? We draw a bunch of
  43. 1:21random lines on the chart and we test
  44. 1:23those random lines against the Fibonacci
  45. 1:26levels. They should look different. What
  46. 1:28I have found overwhelmingly, Fibonacci
  47. 1:31levels are not special. Moving averages
  48. 1:33are not special. VWAP is not special.
  49. 1:35They just should look different. And if
  50. 1:37they don't look different, I don't
  51. 1:39believe that something can be really
  52. 1:41significant in the market and also
  53. 1:43invisible in the data.
  54. 1:44>> In this episode, Adam reveals exactly
  55. 1:47what those tricks are, the one holding
  56. 1:49period where his data says a retail
  57. 1:51trader has the best chance of winning,
  58. 1:53and why stacking six confluences on your
  59. 1:55entry just makes you more confident, not
  60. 1:58more right.
  61. 1:59>> We started off talking
  62. 2:00>> correlation versus causation, signal
  63. 2:02versus noise, the evolution and alpha
  64. 2:03decay and where the world is headed. I
  65. 2:05feel like maybe timeless things would be
  66. 2:07momentum, volatility, trend and
  67. 2:08direction. Would this stand the test of
  68. 2:10time and therefore be better places to
  69. 2:12build an edge around as opposed to when
  70. 2:14you open up a technical analysis book
  71. 2:16and what you may find in there.
  72. 2:17>> I think I have pretty good data on this.
  73. 2:19Um
  74. 2:24>> Adam, so I've been on a US tour right
  75. 2:26now and I'm I want to say pretty
  76. 2:27confused.
  77. 2:28>> Yep.
  78. 2:29>> And the reason for that and I want to
  79. 2:30look for greater truth with yourself. We
  80. 2:31went for dinner last night and you
  81. 2:33strike me as a very introspective and
  82. 2:34philosophical person. And then the
  83. 2:36pursuit for today's conversation is to
  84. 2:38find the truth when I find successful
  85. 2:40traders on both sides of of defense,
  86. 2:43let's say.
  87. 2:44I have traders that tell me, look, the
  88. 2:45market is super complicated because you
  89. 2:47have
  90. 2:48the smartest people in the world with
  91. 2:49the best resources and money and
  92. 2:51everything to like carve out an edge and
  93. 2:52now it's to the point where you have
  94. 2:53Jane Streets executing in nanoseconds
  95. 2:56and it's really difficult and you're
  96. 2:57always subject to alpha decay so you got
  97. 2:58to move fast. Then you have the other
  98. 3:00camp which is also verified made profits
  99. 3:03which is timeless principles, keep it
  100. 3:05simple, a trend line and a doji and a
  101. 3:07trend is enough.
  102. 3:09And there there are opposite sides so
  103. 3:10where would you lie? And
  104. 3:13yeah, what is your beliefs on this topic
  105. 3:14of alpha decay and just the evolution of
  106. 3:17the market or do timeless principles
  107. 3:19edge from the 90s can still exist today?
  108. 3:22>> Both of those things can be true and I
  109. 3:23think they are.
  110. 3:24I think so for
  111. 3:26decades I've structured my trading
  112. 3:28around pullbacks and consolidations and
  113. 3:30trends. And you know, market makes a big
  114. 3:33move and you know, we can take this
  115. 3:35apart
  116. 3:36later, you know, exactly what the forces
  117. 3:38are behind it. Market makes a big move
  118. 3:40and then makes a pause and then
  119. 3:42based on reading how the patterns
  120. 3:44develop, you find the scales tilt a
  121. 3:47little bit in the direction of a
  122. 3:48continued move in that initial
  123. 3:50direction. So that is a fundamental
  124. 3:54principle. And you said from the 90s, I
  125. 3:59would say much deeper, you know, if we
  126. 4:01go back the entire history of recorded
  127. 4:04price data, we go back to we have price
  128. 4:06data from Europe in the Middle Ages. We
  129. 4:09have barley prices from the 1300s.
  130. 4:11Uh we have some of those early clay
  131. 4:14tablets the Phoenicians had were
  132. 4:16accounting tablets and records of prices
  133. 4:18for merchants. And academics have
  134. 4:20reconstructed those prices. And of
  135. 4:23course, you know, no nobody was keeping
  136. 4:24tick data then. They would
  137. 4:27not record daily prices. They would
  138. 4:29record when big changes happened. But
  139. 4:31I've inspected those price series. I
  140. 4:33think that one thing that makes my work
  141. 4:37almost unique is I've looked at the full
  142. 4:40scale of available data in financial
  143. 4:42markets all the way back to the
  144. 4:43Phoenicians. And I would tell you the
  145. 4:46way that I trade looking for a market to
  146. 4:48make a big move, maybe get a little bit
  147. 4:50ahead of itself, maybe consolidate, make
  148. 4:53a pullback or flag or whatever we're
  149. 4:54going to call it, and then make a
  150. 4:56continued move in that direction. That's
  151. 4:58a fundamental principle that has existed
  152. 5:00as long as we've had financial markets.
  153. 5:03On the other hand,
  154. 5:04if I were to fire up my computer at home
  155. 5:08and decide to to compete with Jane
  156. 5:10Street,
  157. 5:11now I have a problem because the length
  158. 5:13of a copper cable, the speed of light
  159. 5:16through that cable matters. And they're
  160. 5:18co-located. I'm at home. I'm playing a
  161. 5:20game where
  162. 5:23I you know, I I brought the wrong tool
  163. 5:25for the job and I can't compete on that
  164. 5:27time frame. So I think both of these
  165. 5:29things are true. And I also think
  166. 5:30Alpha's decay. And I think that the
  167. 5:33specific structure that I built for
  168. 5:35decades around this pullback
  169. 5:38I think the way that I was trading at
  170. 5:4010, 15, 20 years ago I think we're
  171. 5:43starting to see some shifts in the
  172. 5:44statistics. And I don't know
  173. 5:47you know, I I don't know exactly why the
  174. 5:49market is changing, but it does change.
  175. 5:51>> Is therefore for me concludes that it
  176. 5:53matters more what boat you're in not how
  177. 5:55hard you row with the Warren Buffett
  178. 5:56quotes where if I'm trying to compete
  179. 5:58against Jane Street, I'm never going to
  180. 6:00win. But if I go to the other approach,
  181. 6:02I have a better chance. What would be a
  182. 6:04retail independent trader's best chance
  183. 6:06in the markets?
  184. 6:08>> I think I've pretty good data on this.
  185. 6:09Um,
  186. 6:11and not going to be an exciting sexy
  187. 6:14answer. It's not going to be day
  188. 6:15trading. Uh, the best chance
  189. 6:20I think is a time frame where the fewest
  190. 6:22people are trying to extract money.
  191. 6:25The markets, you know, use whatever
  192. 6:27metaphor you want. The markets are a
  193. 6:29battle. They're a war.
  194. 6:31I think they're an ecosystem. I didn't
  195. 6:32come up with that. Andrew Lowe talks
  196. 6:34about markets adapting and thinking
  197. 6:36about different market participants like
  198. 6:37predator and prey, and they evolve and
  199. 6:40they change over time. So if you're
  200. 6:42competing on a very short time frame,
  201. 6:46you're competing with people who have
  202. 6:49infrastructure, data that that you do
  203. 6:52not have. Uh, if you're competing on the
  204. 6:56longer-term time frame, you're competing
  205. 6:58with people that have access to
  206. 7:00fundamental data. If you think you're
  207. 7:01doing fundamentals at home, well, you're
  208. 7:03up against firms who are sending
  209. 7:06analysts to the companies. You know,
  210. 7:07they're spending time in the factories.
  211. 7:09They're talking to management. They know
  212. 7:11what whatever you think you know, you
  213. 7:13don't know what they know. And And you
  214. 7:15don't have access to that information.
  215. 7:17If you think you're trading oil on
  216. 7:19fundamentals, maybe you are, but do you
  217. 7:21have access to the data feeds that show
  218. 7:23you where the tankers are in the ocean,
  219. 7:25how full they are, when they're Do you
  220. 7:26Do you
  221. 7:27Is Is that the game you're going to
  222. 7:29compete as a trader at home? Probably
  223. 7:31not.
  224. 7:32I think a simple concrete answer is
  225. 7:34looking for holds that are probably a
  226. 7:36little bit longer than you're going to
  227. 7:37be inclined for, so not day trades. You
  228. 7:40do need that overnight risk, but
  229. 7:41somewhere in the, you know, let's say
  230. 7:43average holding period two to five days
  231. 7:45is probably the sweet spot for a retail
  232. 7:47trader.
  233. 7:47>> [clears throat]
  234. 7:48>> You mentioned how there are certain
  235. 7:49principles that you could trace back to
  236. 7:51hundreds of years.
  237. 7:52>> Yep.
  238. 7:52>> And I would suppose that common
  239. 7:54denominator between then and now is
  240. 7:56human emotion.
  241. 7:57Um,
  242. 7:58>> Yeah.
  243. 7:58>> greed, optimism, euphoria, then despair,
  244. 8:01XYZ, which makes total sense and that
  245. 8:03would be reflected naturally in price
  246. 8:05action until you go digital and until
  247. 8:07you go algorithmic. And now, if I'm not
  248. 8:09wrong, 85 or 90% of price movements or
  249. 8:13orders are algorithmically generated or
  250. 8:16executed.
  251. 8:17When I see the comments online, it's
  252. 8:19like, well, yeah, but humans coded it,
  253. 8:20so it's the emotions are baked in. I'm
  254. 8:22not quite sure, but nonetheless, does
  255. 8:24human emotion reflect when the majority
  256. 8:26of the market is algorithmic?
  257. 8:28>> So far, yes. So, you know, we're
  258. 8:30recording at a point in time and the
  259. 8:33markets may evolve very quickly and I
  260. 8:35may say things today that look very
  261. 8:37foolish 5 years from now. So, you know,
  262. 8:39I will reserve the right to say that I
  263. 8:40was wrong and sure, I'm going to be
  264. 8:42wrong in some things, but I think so. I
  265. 8:44think this argument that people make
  266. 8:46that
  267. 8:47algorithms So first of all, many
  268. 8:50algorithms are are execution algorithms.
  269. 8:52That's that's all they are. You know,
  270. 8:54people think like there's some algo that
  271. 8:56determines price. No, you know, like if
  272. 8:59if I and you know, I've been this trader
  273. 9:01where if I need to execute
  274. 9:04a lot relative to the volume in a
  275. 9:06market, I will either farm it to a
  276. 9:08trader who can do it algorithmically or
  277. 9:10to an algo. And that algo is going to
  278. 9:12you know, it's going to be benchmarked
  279. 9:13against VWAP over the time period. But
  280. 9:16it's the algorithm is just figuring out
  281. 9:18how to nibble my order without really
  282. 9:20moving the market because you know, if I
  283. 9:21push the button, the the market moves.
  284. 9:24Many algorithms are just that and what
  285. 9:27they're looking to do is buy when the
  286. 9:31market is and you know, we can flip for
  287. 9:33selling, but you know, they're they're
  288. 9:34looking to extract some minimal value
  289. 9:37and it's not something that's completely
  290. 9:40disconnected from the way that humans
  291. 9:43trade, the way that humans think. And
  292. 9:45the opportunistic algorithms, if you
  293. 9:47will, well, what are they looking for?
  294. 9:50They're looking for in most cases
  295. 9:51extremes and some kind of mean
  296. 9:53reversion. That's by and large. So what
  297. 9:56are they looking for? How do we get to
  298. 9:58those extremes?
  299. 9:59Either error or emotion, right? And in
  300. 10:02financial markets, maybe emotion is
  301. 10:03error. I don't know. Um
  302. 10:06so I
  303. 10:08we are seeing I'm seeing in my
  304. 10:11quantitative work,
  305. 10:12uh alphas decaying. I'm seeing the
  306. 10:14markets change in some fundamental ways.
  307. 10:16It may be exactly what you're asking
  308. 10:18about. It may be
  309. 10:20new kinds of AI driving market action,
  310. 10:24but I think over the, you know, as the
  311. 10:26world's been more and more digital going
  312. 10:28back to the late 80s,
  313. 10:30I think at least up to this point where
  314. 10:32we're having this conversation, I think
  315. 10:34the idea of emotion is still valid.
  316. 10:36>> It's interesting. You mentioned before
  317. 10:39if we reached out to you and said, hey,
  318. 10:41let's do a podcast, you would have said
  319. 10:43no because I've got a system. It's
  320. 10:44working. I'm making money and that would
  321. 10:45be a boring conversation.
  322. 10:46>> not that interesting, yeah.
  323. 10:47>> Which sounds more interesting to me, but
  324. 10:49now we're at a time where you're saying,
  325. 10:51well, it seems like we're at an
  326. 10:52inflection point in history where you're
  327. 10:53sensing change, you're sensing decay,
  328. 10:55and you're seeing the emergence of
  329. 10:57quantitative power that maybe reshaping
  330. 10:59the landscape and you're saying my
  331. 11:00beliefs today might look foolish in five
  332. 11:02years. Are we at a
  333. 11:05inflection point in history right now in
  334. 11:06the markets?
  335. 11:07>> I believe we are at an inflection point
  336. 11:09in history.
  337. 11:10Full stop.
  338. 11:13I suspect,
  339. 11:14you know, in my lifetime I saw the
  340. 11:17arrival of the internet which changed
  341. 11:18everything, the internet and email. And
  342. 11:21I know generations before us saw
  343. 11:23electricity, industrial revolution.
  344. 11:25We've seen these major inflection
  345. 11:28points. I suspect today where it's
  346. 11:31something that
  347. 11:32we haven't like this is maybe
  348. 11:34invention of writing. This is maybe
  349. 11:36agriculture. This this could be one of
  350. 11:39those civilization level inflections.
  351. 11:42And I think there are concerns outside
  352. 11:45of I have many concerns outside of
  353. 11:47financial markets, how we navigate this
  354. 11:49as a species and as a culture and and as
  355. 11:52humans.
  356. 11:53>> With the objective I guess today of
  357. 11:55understanding truths of the market, I'd
  358. 11:57like to know where I should begin
  359. 11:58because it seems like you've had a
  360. 11:59journey in your career of styles of
  361. 12:01trading, what you use from day trading
  362. 12:03M1 to where you are now.
  363. 12:06A lot of traders begin at strategies and
  364. 12:08and that would be like broad schools of
  365. 12:09thought whereas Elliot wave theory,
  366. 12:11Wyckoff theory, smart money concepts,
  367. 12:13and Fibonacci support resistance, and
  368. 12:15then you'd try and
  369. 12:17put that to fit into the market. The
  370. 12:19other approach would be the I guess more
  371. 12:21algorithmic approach, which is
  372. 12:23past price could predict future price,
  373. 12:25and as a result, I'm just going to go
  374. 12:27through data to find things.
  375. 12:29Most people start at the other side,
  376. 12:31which is schools of thought because it
  377. 12:33feels more approachable.
  378. 12:34But nonetheless, the ones that do or
  379. 12:36don't work When I think of let's say a
  380. 12:38Fibonacci,
  381. 12:39is there something magical about a
  382. 12:40Fibonacci level that ends up as as a
  383. 12:43rejection point, or is it kind of like a
  384. 12:45bell curve and you have these areas of
  385. 12:47increased probability? I don't know.
  386. 12:49But what I'm more concerned myself with
  387. 12:51is this a correlation or causation? Does
  388. 12:53the market reverse because of the
  389. 12:55Fibonacci or this nicely reflects a
  390. 12:57reversal point? If I could start to
  391. 12:59think in first principles and start to
  392. 13:00think in causation,
  393. 13:02where would I begin?
  394. 13:04>> So
  395. 13:06the way you framed that, this is
  396. 13:08another way to rephrase this is the
  397. 13:10self-fulfilling prophecy. Is the
  398. 13:12Fibonacci level, moving average, VWAP,
  399. 13:15whatever we're talking about, a
  400. 13:17self-fulfilling prophecy?
  401. 13:19My answer to that I do care deeply about
  402. 13:22the truth. We're going to come back to
  403. 13:23that. But my answer to is it a
  404. 13:25self-fulfilling prophecy is
  405. 13:27I don't think I care. What I want to
  406. 13:30know first of all is and let's talk
  407. 13:32about Fibonacci specifically,
  408. 13:34I want to know
  409. 13:36does the market reverse at that level,
  410. 13:37first of all? That's the question.
  411. 13:40So how would we test that?
  412. 13:44How how it it it's not so easy. How how
  413. 13:46would we structure a test?
  414. 13:47>> my first concern would be how do I
  415. 13:49separate let's say a broken clock where
  416. 13:51my watch is stuck at 6:00. It's correct
  417. 13:52twice a day, 6:00 and 6:00 p.m. The rest
  418. 13:54of the time is wrong. But the times is
  419. 13:56correct. If I give if I attribute
  420. 13:58meaning to it, I'm attributing
  421. 14:00meaning to noise. So I'd have to
  422. 14:02separate signal from noise.
  423. 14:05I'll pass it back to you and help me
  424. 14:06navigate that one.
  425. 14:07>> Well, there's also And then there's also
  426. 14:10the
  427. 14:11cognitive issue that you're going to
  428. 14:13remember things that look and feel
  429. 14:15significant. You know, if if you look at
  430. 14:17your watch and it says 6:00 00 and then
  431. 14:20you look at it again, it says 7:00 and
  432. 14:22you're like, oh, you know, I looked at
  433. 14:24my clock. So you and you know that's
  434. 14:26meaningless, but it's a it's a data
  435. 14:28point that jumps out. Uh so the question
  436. 14:31is
  437. 14:34if I understand what you're saying, I
  438. 14:36have a Fibonacci level.
  439. 14:38First of all, let's define what we think
  440. 14:40should happen. Price should reverse more
  441. 14:42from the Fibonacci level. Is that is
  442. 14:43that what we're saying? Uh because we do
  443. 14:47kind of need to think around the corners
  444. 14:49because that's one possibility. Another
  445. 14:51possibility is maybe price is drawn to
  446. 14:53the Fibonacci level. Maybe price engages
  447. 14:57like you know, that could be another way
  448. 14:58to think about it. Uh but regardless,
  449. 15:00both of those
  450. 15:02your question is how do I know it's not
  451. 15:03the stopped clock being right twice a
  452. 15:05day, right? Um there's a very simple
  453. 15:08technique for that
  454. 15:10which is we draw a bunch of random lines
  455. 15:12on the chart
  456. 15:13and we test those random lines against
  457. 15:16the Fibonacci levels. And this is a
  458. 15:20you know, it's funny. Uh
  459. 15:22I didn't have a quantitative background
  460. 15:24and I didn't know how to test any of
  461. 15:27this stuff. When I when I started
  462. 15:29trading out of college, I was a
  463. 15:30musician. I had no math classes in my
  464. 15:33pocket. Uh so I had to teach myself
  465. 15:35everything. And then
  466. 15:37you know, initially you get very excited
  467. 15:40because you see things that make money.
  468. 15:41Uh for instance, it's quite easy to
  469. 15:44develop a system that buys stocks that
  470. 15:46makes money because stocks on average go
  471. 15:49up over time, right? So in fact, if you
  472. 15:51develop a system where you're buying
  473. 15:53stocks and doesn't make money
  474. 15:55that may be very interesting to me today
  475. 15:57because maybe you found an edge. But the
  476. 15:59question is, am I just getting the index
  477. 16:01drift or am or am I getting am I getting
  478. 16:04alpha you know,
  479. 16:05we we know what we're talking about
  480. 16:06here. Are we getting alpha? So what you
  481. 16:08can do is test against these dummy
  482. 16:11levels. And I wouldn't even you know,
  483. 16:14with the Fibonacci level
  484. 16:16I wouldn't even say
  485. 16:19I wouldn't even define the bounce. It
  486. 16:20would be enough for me initially to say,
  487. 16:22I have my interactions with Fibonacci
  488. 16:25levels. And what happens after those. I
  489. 16:27would go through and I would look every
  490. 16:29time we touch a level and then, you
  491. 16:32know, you you ask me, well, do you
  492. 16:33really need to touch it? And I say,
  493. 16:34okay, we can put a little band around
  494. 16:36it, whatever. Every time we get close to
  495. 16:38a level,
  496. 16:39let's look what happens going forward.
  497. 16:42And then because we're doing a lot of
  498. 16:43tests, let's also test exact engagements
  499. 16:45and we have a bunch of levels and you
  500. 16:47know, should we test 50%? It's not
  501. 16:49really a Fibonacci level, but the
  502. 16:50Fibonacci traders use it. Uh we should
  503. 16:52talk about Fibonacci and sacred
  504. 16:53geometry. It's a, you know, entirely
  505. 16:56separate topic. But sticking on the
  506. 16:57tests now,
  507. 16:59so we look at what happens. We look at
  508. 17:01the outcomes from those engagements and
  509. 17:03then we look at the outcomes from the
  510. 17:04random lines. And they should look
  511. 17:08different.
  512. 17:09And of course, you know, one one of the
  513. 17:11we have to use the right kind of
  514. 17:13statistics because if you look closely
  515. 17:14enough, you'll see differences where
  516. 17:16there aren't differences. So we have to
  517. 17:18understand the variability. But what I
  518. 17:20have found overwhelmingly, and this is
  519. 17:24this is where people probably start
  520. 17:25throwing things at the screen or are
  521. 17:27spiking their phone on the ground,
  522. 17:30all of these things that people think
  523. 17:32were Fibonacci levels are not special.
  524. 17:35Moving averages are not special. VWAP is
  525. 17:38not special. Simple test of a moving
  526. 17:40average, by the way, you want to tell me
  527. 17:41a 50 period or 200 period moving average
  528. 17:44is special. Fine. Let's test your 200
  529. 17:46period. Test it against 210 and 109, you
  530. 17:49know. And again, they just should look
  531. 17:51different. And if they don't look
  532. 17:53different,
  533. 17:55I don't believe that something can be
  534. 17:57really significant in the market and
  535. 17:59also invisible in the data. It might be
  536. 18:02hard to see. It might be subtle. It
  537. 18:03might be slight and it might not show
  538. 18:06up. You know, I don't know. Like you you
  539. 18:09told me you think price bounces off a
  540. 18:10Fibonacci level. Maybe it does. Maybe it
  541. 18:13accelerates through a level. Our tests
  542. 18:14would have found that. Maybe after the
  543. 18:17level volatility gets smaller. Our tests
  544. 18:19would have seen that because we're
  545. 18:20looking all we're saying is, you know,
  546. 18:22we're saying I don't know. I'm very
  547. 18:24naive. I have these two sets of
  548. 18:26distributions, which are the, you know,
  549. 18:28forward distributions. Let me look and
  550. 18:30see if they look different. So that's
  551. 18:31the test.
  552. 18:33>> I love to express this conversation
  553. 18:34towards It seems like you've done this
  554. 18:36test that you're describing today on
  555. 18:37Fibonacci and you probably done it on
  556. 18:39everything else, trend lines and support
  557. 18:41resistance, XYZ.
  558. 18:42What What conclusions did you derive?
  559. 18:44>> Uh there are no with very few
  560. 18:47exceptions. And you want to know I I'll
  561. 18:49tell you where some of them are. With
  562. 18:51very few exceptions, there are no
  563. 18:52special levels. Uh in general, the from
  564. 18:56a quantitative perspective, the and it
  565. 18:59it varies according to asset class a
  566. 19:01little bit. Uh for instance, stock
  567. 19:04indexes have a natural tendency for mean
  568. 19:07reversion.
  569. 19:08All markets have mean reversion. Stock
  570. 19:09indexes tend to have a little bit more.
  571. 19:12Interest rate sensitive instruments, FX,
  572. 19:15you know, bond futures, uh
  573. 19:17less mean reversion, more continuation.
  574. 19:20So if I had to if if you told me uh a
  575. 19:23market went down a lot today
  576. 19:27and you're going to buy it on the close,
  577. 19:28if you tell me it's a stock index
  578. 19:30or it's a currency, you have a different
  579. 19:33probability of having a profitable
  580. 19:34trade. You have much better chance of
  581. 19:35making money on the stock index. So one
  582. 19:38thing we will see is if you're looking
  583. 19:40at moves down to moving averages in
  584. 19:42stock indexes, they naturally bounce.
  585. 19:45But guess what? So do our random lines,
  586. 19:47too. Uh you also can I've run tests
  587. 19:51where, you know, I said we should test
  588. 19:53the 200 against the 210 and the 190. Uh
  589. 19:57we also should test it against a move
  590. 19:59How about a moving average that changes
  591. 20:01its period every every bar?
  592. 20:04So today it's a 90 period. Tomorrow it's
  593. 20:06130 130 period. So it's it's a
  594. 20:09meaningless line on the chart.
  595. 20:11>> This is where I get confused because I
  596. 20:13have conversation with certain
  597. 20:14individuals, including recently, where
  598. 20:16he broke down elements of his edge. He
  599. 20:18was like, yeah, one of my models is like
  600. 20:20once we're above the the 200 moving
  601. 20:22average, if we have three down days, uh
  602. 20:24going to do X. I'm like, anything else?
  603. 20:27And they're like, no, that's it. And
  604. 20:28then I speak to another person. He's
  605. 20:29like, yeah, I just wait for a EMA cross.
  606. 20:32I wait for a trend line top or I wait
  607. 20:34for a breakout. I was like, oh, do you
  608. 20:35wait for the test? Do you use a
  609. 20:36Fibonacci? Do you use a Do you like No,
  610. 20:38just a breakout. I was like, anything
  611. 20:39more? Nothing. And I And then I have the
  612. 20:41other side of the camp and we have
  613. 20:42really deep technical conversations down
  614. 20:44to M1, down to the philosophical side
  615. 20:46that we're touching upon. How can these
  616. 20:48both coexist? Which is back to the
  617. 20:50original question.
  618. 20:50>> Okay, they can both coexist in a few
  619. 20:52ways. So
  620. 20:54going back to the general behavior of
  621. 20:56stock markets, when stock markets are
  622. 20:58significantly off their highs,
  623. 21:01which would put you under the 200 day
  624. 21:02moving average. And the 200 day moving
  625. 21:04average is not a magical line in the
  626. 21:05sand, but in general
  627. 21:07with stock indexes or individual stocks,
  628. 21:10if you're under the 200 day moving
  629. 21:11average, you're going to see a different
  630. 21:13kind of behavior. You're going to see a
  631. 21:15um
  632. 21:17in the more recent data, you see more
  633. 21:18volatility.
  634. 21:20And this again, this is not true in
  635. 21:22currencies because there's not really an
  636. 21:24up and down. It depends, you know, what
  637. 21:26what you're weighing it against. Um
  638. 21:29but with stocks,
  639. 21:31under the 200 day moving average, you
  640. 21:33will see different kinds of behavior.
  641. 21:36Doesn't It doesn't mean the 200 day
  642. 21:37moving average itself is magical, but it
  643. 21:39does mean that when you're off of that
  644. 21:41high and maybe uh
  645. 21:43you know, if you're using in a
  646. 21:45systematic approach, you're using the
  647. 21:47200 day moving average. Well, you have
  648. 21:48to use something. And you're capturing
  649. 21:51this aspect of behavior. Maybe it could
  650. 21:52be a 200 day median, which is actually
  651. 21:55quite you know, computationally is a
  652. 21:57little bit more expensive to compute
  653. 21:59every day. Um maybe you could just look
  654. 22:01at where you are in the yearly range or
  655. 22:03something like that or how far you
  656. 22:04You're going to get a similar effect.
  657. 22:06The
  658. 22:07>> Th- This is down to the self-fulfilling
  659. 22:08idea that there's nothing magical about
  660. 22:10the moving average, but generally people
  661. 22:12would be behaving in a certain way below
  662. 22:13it.
  663. 22:14>> It's not It's not even that, you know, I
  664. 22:15I would be perfectly happy. Our test of
  665. 22:17Fibonacci levels, which I promise you,
  666. 22:21shows no difference with the with the
  667. 22:23Fibonacci level. If they were
  668. 22:25self-fulfilling prophecy, we would see
  669. 22:28the difference. We would see the action
  670. 22:29there. And then I would be quite
  671. 22:31comfortable telling you,
  672. 22:32I don't really care why it's I don't I I
  673. 22:35can't know why it's doing that, but it
  674. 22:37does that. And I see it in the data. I
  675. 22:39don't even see it in the data. If the
  676. 22:41200-day if funds were buying at the
  677. 22:4450-day or the 200-day or you know the
  678. 22:46nonsense that that people say, you would
  679. 22:48see it at those averages. You do not see
  680. 22:51it at those averages. Now that can be
  681. 22:52true and it can also be true that when
  682. 22:55stock indexes are significantly off
  683. 22:57their highs, you see a different kind of
  684. 22:58behavior primarily with respect to
  685. 23:00volatility. I don't think there is a
  686. 23:04strong directional edge,
  687. 23:07but there is you you you'll have more
  688. 23:10volatile events under the historically
  689. 23:12at least under the 200-day moving
  690. 23:13average.
  691. 23:13>> what I personally done I'm with you. I'm
  692. 23:15on the same page and I like to avoid
  693. 23:17looking at things that are correlated
  694. 23:19like a head and shoulder pattern. I
  695. 23:20don't know what that means and it could
  696. 23:21work but it is signal versus noise. I
  697. 23:23don't know if it's statistically
  698. 23:24relevant. Same for Fibonacci, same for
  699. 23:25trend line, support resistance, X Y, the
  700. 23:26list goes on.
  701. 23:28But I do like to think in in terms of
  702. 23:30causation. What could cause the market
  703. 23:32to move? And I find it
  704. 23:34as as as time goes on or as I'm
  705. 23:36projecting into further time horizons, I
  706. 23:39guess the crystal ball gets a bit hazy.
  707. 23:40So I think it becomes harder to predict.
  708. 23:42Naturally as a result of that I'm doing
  709. 23:44these days of the last decade what you
  710. 23:46were initially doing and I moved away
  711. 23:48from which is I like to be on the M 1. I
  712. 23:49like to be on shorter time horizons
  713. 23:51because I think the causation
  714. 23:53edge exists there. And what in the
  715. 23:56language I think of is supply and demand
  716. 23:58surely has to be causative. Market
  717. 24:00impact of large institutions can create
  718. 24:02a small causative event. Now I don't
  719. 24:04know if that's going to lead to a full
  720. 24:06move, but it might give a reaction. So I
  721. 24:08like to think in these terms. Am I on a
  722. 24:10better line than or a better direction
  723. 24:12in the industry than someone who is
  724. 24:13looking at Fibonacci and trend lines and
  725. 24:15these generic things?
  726. 24:16>> I think so. I mean if you're trading on
  727. 24:18the 1-minute chart,
  728. 24:20one of the other things that will test
  729. 24:22out well which you're probably aware of,
  730. 24:24previous day's high and low,
  731. 24:26that's a real thing.
  732. 24:27And if we
  733. 24:30you can test this very easily. You can
  734. 24:32do the test we were talking about, look
  735. 24:33at engagements of the high and low,
  736. 24:36look at random lines, throw your
  737. 24:37Fibonacci levels in there, and what
  738. 24:39you're going to see is that one of these
  739. 24:41things is different than the other. The
  740. 24:43high and low looks different than the
  741. 24:45Fibonacci and looks, you know,
  742. 24:47it's non-random behavior there. So I I
  743. 24:51think you're correct there. I, you know,
  744. 24:53as humans we we tell stories, right?
  745. 24:55We're we're we're we're our
  746. 24:58culture is tied all of our cultures are
  747. 25:02tied together by the stories we tell.
  748. 25:04And we tell these stories about
  749. 25:05ourselves and we're telling stories here
  750. 25:06today.
  751. 25:08And there's something that's very
  752. 25:09appealing about that. And as somebody
  753. 25:11who's done a decent amount of spiritual
  754. 25:14exploration, stories are important to
  755. 25:16me, but
  756. 25:17as a
  757. 25:19quantitative trader, I know the market
  758. 25:22is, you know, the story I would tell you
  759. 25:24about the market is the market is
  760. 25:26very fickle. The market the the market
  761. 25:28can change
  762. 25:30can can be
  763. 25:32making me feel one way and then it can
  764. 25:35change very quickly. So I know that I
  765. 25:36can't get very attached to my stories.
  766. 25:39I
  767. 25:42you know, there's also at any point in
  768. 25:43the market there's so many different
  769. 25:45influences and different people trading
  770. 25:47that I don't think I have
  771. 25:50you know, in
  772. 25:51in full humility I don't think I I can
  773. 25:53understand those influences, but I will
  774. 25:55also tell you I find the Wyckoff
  775. 25:58framework very useful. I know it's naive
  776. 26:01and
  777. 26:02the viewers
  778. 26:04where you think about these big buyers
  779. 26:06and big sellers driving the market. They
  780. 26:08may not even be real, but you can think
  781. 26:10what are buyers trying to do at this
  782. 26:12point? What are sellers trying to do at
  783. 26:14this point? And if you use those
  784. 26:16frameworks, even if these stories are
  785. 26:18completely fictitious, we're still
  786. 26:21saying something about
  787. 26:23the reality that has some any of the
  788. 26:26story can be useful. Like when
  789. 26:29when people in a previous culture told a
  790. 26:31story about a goddess that went into the
  791. 26:34underworld so all the plants died and
  792. 26:36the winter and then you would know when
  793. 26:37to replant.
  794. 26:39I don't think they thought that was real
  795. 26:40or not but it guided their decisions for
  796. 26:43how how they planned to get through the
  797. 26:45winter and when the spring came and I
  798. 26:48think this this goes back to the
  799. 26:49question you asked before should you
  800. 26:51start where should you start should you
  801. 26:53start analyzing data quantitatively or
  802. 26:56should you start kind of trying to wrap
  803. 26:58your arms around the market.
  804. 27:00I would say start trying to wrap your
  805. 27:02arms around the market but you know that
  806. 27:03that's my bias because I think at heart
  807. 27:07despite all the quantitative work I'm an
  808. 27:09artist I'm a musician you know I'm I'm a
  809. 27:11weird person because
  810. 27:13this quantitative side but I also have
  811. 27:16this holistic artistic approach and I
  812. 27:19think study study patterns
  813. 27:24trade you know like
  814. 27:25you learn don't lose much money keep
  815. 27:28keep your losses small but you can do
  816. 27:30all of the theoretical trading you know
  817. 27:33reading and back testing and it all
  818. 27:36changes when you get in the market and
  819. 27:37the market punches you in the face and
  820. 27:39you oh okay they there's something
  821. 27:41there.
  822. 27:41>> You mentioned that that Wyckoff could be
  823. 27:43a little bit naive but even looking at
  824. 27:45it comparing it to the other things
  825. 27:47we've spoken about that one still makes
  826. 27:48sense to me maybe not the stories maybe
  827. 27:50not the labels and the the UTAD and the
  828. 27:53LPSY and all these let's say acronyms we
  829. 27:55give to it but the philosophy of
  830. 27:58institutional buyers loading up that's a
  831. 28:00cause it's like an elastic band I'm
  832. 28:01pulling it back if I let go surely it
  833. 28:03should
  834. 28:04it's a cause and effect relationship so
  835. 28:06that makes more sense to me.
  836. 28:07>> What are they trying to do in the
  837. 28:09deception in the market?
  838. 28:10>> naturally there's going to be a lot of
  839. 28:12nuance to it like not every buyer is
  840. 28:14equal someone is a speculator someone is
  841. 28:15a utilizer of the currency some people
  842. 28:18is trying to hold for long term someone
  843. 28:19short term someone's weak handed XYZ so
  844. 28:21it does get nuanced of course but the
  845. 28:22general principle of I want to trade in
  846. 28:25the market from causative perspectives
  847. 28:27accumulation fuels causative what other
  848. 28:29things could fuel causative and again
  849. 28:32avoiding just storytelling which could
  850. 28:34be guiding but I don't want to get into
  851. 28:36that because that can um yeah, it can it
  852. 28:38can just get a bit weird.
  853. 28:38>> podcast, right? That's because it's a
  854. 28:39little weird.
  855. 28:41Um
  856. 28:42I think this framework is very useful
  857. 28:43too. And I think, particularly as a
  858. 28:46discretionary trader,
  859. 28:48when you
  860. 28:50you do all of the analysis and it still
  861. 28:52comes down to you making a decision. And
  862. 28:55I find that my decisions are best
  863. 28:58motivated by understanding what I think
  864. 29:00these groups of people are trying to do.
  865. 29:02If I think I see, you know, okay, buyers
  866. 29:04are
  867. 29:06you said accumulating here, and I think
  868. 29:07that's a word that in some way, you
  869. 29:09know, it's overused, but you know, if I
  870. 29:11keep seeing that the market keeps
  871. 29:12getting depressed on this time frame uh
  872. 29:15like there's buying there. There's some
  873. 29:16and
  874. 29:17how much do I trust it? I don't know. I
  875. 29:19can't really trust anything that much,
  876. 29:21but it's something that is happening.
  877. 29:22There's somebody there that is buying.
  878. 29:24And then like you said, you know, what
  879. 29:26happens when you move into you move
  880. 29:28above a resistance level um
  881. 29:31and all of a sudden there's not the
  882. 29:32selling pressure to meet that and then
  883. 29:34the market moves. And you saw it because
  884. 29:36you saw the positioning beforehand. Or,
  885. 29:38you know, if you're living in an
  886. 29:40intraday world, you know, you see you
  887. 29:42keep coming back to the high of the day
  888. 29:45and you can't really back off from it
  889. 29:47and you go dull and dead and quiet at
  890. 29:49the high of the day, that's telling you
  891. 29:51something. You know, the they're they're
  892. 29:53not letting it back off much. So buyers
  893. 29:55I think that
  894. 29:57whether or not this is actually a thing,
  895. 29:59it is for a discretionary trader a
  896. 30:00powerful tool for understanding price
  897. 30:02action.
  898. 30:03>> We started off talking about correlation
  899. 30:04versus causation, storytelling versus
  900. 30:06signal versus noise, and then also with
  901. 30:09the evolution and alpha decay and where
  902. 30:11the world is headed.
  903. 30:12I feel like maybe timeless things would
  904. 30:14be momentum, uh volatility, uh trend and
  905. 30:17direction. These I suppose would always
  906. 30:19be baked in.
  907. 30:19>> Add mean reversion.
  908. 30:21>> Add mean reversion, yes. Would would
  909. 30:22this stand the test of time and
  910. 30:24therefore be better places to build an
  911. 30:26edge around as opposed to when you open
  912. 30:28up a technical analysis book and what
  913. 30:29you may find in there.
  914. 30:30>> I think so, but I think if you open up a
  915. 30:33good tech and you know, not to talk my
  916. 30:35book, but if you up my book, my book is
  917. 30:37going to be centered around momentum and
  918. 30:39mean reversion, and the patterns are an
  919. 30:41expression of that. So I think, you
  920. 30:44know,
  921. 30:45barring our weird like
  922. 30:47civilizational financial market ending
  923. 30:49discussion earlier, which was pretty
  924. 30:51weird, as long as there are functioning
  925. 30:53financial markets, these forces of mean
  926. 30:55reversion and momentum will continue to
  927. 30:57play. I can't imagine financial markets
  928. 30:59where they will not play out, but they
  929. 31:01may play out in ways that mean the
  930. 31:03specific patterns, the ways we trade
  931. 31:05them don't necessarily work. So we may
  932. 31:07need but
  933. 31:09I don't
  934. 31:10and there are some other principles,
  935. 31:12right? We should talk about options
  936. 31:13traders doing things with volatility,
  937. 31:15doing things with time. There are things
  938. 31:17with relative value. You know, some
  939. 31:18people will tell you that the only thing
  940. 31:20that matters is price, you know, only
  941. 31:22price pays. But there is there are other
  942. 31:24ways to trade, right? You can set up
  943. 31:26volatility structures.
  944. 31:28You
  945. 31:30these things may
  946. 31:32it's kind of a level of sophistication
  947. 31:34up from trading chart patterns.
  948. 31:37It's going to be there some way, you
  949. 31:39know, and we may
  950. 31:41I I could imagine a world where the
  951. 31:43chart patterns are so heavily arbed, but
  952. 31:46they don't arb the volatility out of
  953. 31:48relative relationships. So, you know,
  954. 31:50there's something to be done with
  955. 31:51spreads perhaps.
  956. 31:54>> We spoke about dinner last night that
  957. 31:56the topic of the industry or the
  958. 31:58buzzword in the industry right now seems
  959. 32:00to be this year is all about order flow
  960. 32:02in the retail lens. A lot of influencers
  961. 32:04are popping up talking about it and it
  962. 32:05just seems to be everyone's looking for
  963. 32:07holy grail, mine's not working, let me
  964. 32:09strategy hop and and try this one out.
  965. 32:11However, it doesn't feel like the other
  966. 32:13ones in the past that I've seen where
  967. 32:14it's just storytelling and something
  968. 32:15repackaged. It feels like something that
  969. 32:17could be causal where
  970. 32:19I'm looking for big orders to come in.
  971. 32:21This could be institutional for prints
  972. 32:23and naturally self-fulfilling. If a
  973. 32:24large institution steps in, price should
  974. 32:26at least move a bit in that direction.
  975. 32:28Is there any merit to
  976. 32:30linking price action to order flow
  977. 32:32behind it?
  978. 32:33>> I think so, yes, but
  979. 32:35where I would
  980. 32:36where where I would take issue with this
  981. 32:38is with it being new, because this is
  982. 32:40this is something, you know, this is
  983. 32:42this is not new.
  984. 32:43Uh, volume profile, I
  985. 32:46I was kind of in the second generation
  986. 32:49of that. It If I understand correctly,
  987. 32:51it was first pitched to the floor
  988. 32:53traders in Chicago who weren't really
  989. 32:55able to make much with it. So then it
  990. 32:57moved to a retail marketing. It's been
  991. 33:00repackaged in, you know, various ways.
  992. 33:02And then order flow, people have been
  993. 33:03looking at order flow ever since markets
  994. 33:05went electronic. And surely there's
  995. 33:07information there. Uh, I think people do
  996. 33:11things with order flow that's probably
  997. 33:12not appropriate and correct. You know,
  998. 33:14the idea that uh
  999. 33:17you see somebody
  1000. 33:19made a big transaction at a price. Does
  1001. 33:22that mean they're going to defend it
  1002. 33:23when it comes back? And I You know,
  1003. 33:24that's one It's one of many ways that
  1004. 33:26people think about it. So
  1005. 33:28I
  1006. 33:30I think particularly for scalpers,
  1007. 33:32there's a lot of information there. But
  1008. 33:34there's also a tremendous amount of
  1009. 33:36deception on the tape. You know, you
  1010. 33:38don't see the displayed liquidity. And I
  1011. 33:40know because I've been I've been on both
  1012. 33:42sides of that. I've been reading the
  1013. 33:43tape and I've also been the guy trying
  1014. 33:46to screw the people reading the tape
  1015. 33:47when I'm had to execute a lot of volume
  1016. 33:50in thin stocks. You know, I I I I know
  1017. 33:51the tricks to do that.
  1018. 33:53>> What does that mean, the second part?
  1019. 33:54>> Well, it means that So say say there's a
  1020. 33:56very Just for instance, there's a very
  1021. 33:58thin stock or it could be an option. It
  1022. 34:00could be a currency at a quiet time of
  1023. 34:02the day and there's not much liquidity.
  1024. 34:03The bid ask is very wide. It's not
  1025. 34:05trading very much and you need to do a
  1026. 34:09big order right now. If you just go and
  1027. 34:12start clearing the book, you're going to
  1028. 34:13move it a lot. And instead, what you
  1029. 34:16might do There's all kinds of tricks.
  1030. 34:18You If you have to buy, you know,
  1031. 34:20whatever quantity Let's say you have to
  1032. 34:21buy 20,000, you don't have to put 20,000
  1033. 34:24in the book and let everybody see, hey,
  1034. 34:26you know, uh you you can put in 100, you
  1035. 34:29can put in 200. You can show that. You
  1036. 34:30can ice break the rest. You can Uh,
  1037. 34:33if you're looking to buy, you might put
  1038. 34:35an order in to sell. You might put a big
  1039. 34:37order in to sell way above the market
  1040. 34:39and then people reading the tape will
  1041. 34:41think, oh, there's actually selling
  1042. 34:42pressure there. Uh,
  1043. 34:44you're you know, you're I I got it
  1044. 34:46flipped, but you know, uh,
  1045. 34:48the
  1046. 34:50so a lot of what you see on the tape is
  1047. 34:53by nature deceptive. I don't think you
  1048. 34:55can take it. And then there's of course
  1049. 34:57the argument for looking at the
  1050. 34:59difference between the display, what you
  1051. 35:01see displayed and what you see actually
  1052. 35:02transacted, that you know, the
  1053. 35:04transactions are a level of reality that
  1054. 35:07the displayed theoretical liquidity is
  1055. 35:10not. There's something there. Uh, I
  1056. 35:13I think that the way that I read charts
  1057. 35:16is
  1058. 35:18effectively a kind of tape reading on
  1059. 35:20the chart.
  1060. 35:21You know, and I I think you probably if
  1061. 35:23you're reading a one-minute chart, I
  1062. 35:25think for instance go back to that
  1063. 35:26situation where you're up against the
  1064. 35:27high of the day and imagine you can't
  1065. 35:30really back off from the high of the day
  1066. 35:32and a few times that it's sold off, it
  1067. 35:34leaves long long wicks, long tails.
  1068. 35:37You know that those were probably just
  1069. 35:40quick probes down. So you're reading the
  1070. 35:43tape as you're reading the chart. And
  1071. 35:45the argument the tape readers will make
  1072. 35:46well, you know, the chart's static,
  1073. 35:48you're not seeing what's actually on the
  1074. 35:50tape, but it's leaving the footprint in
  1075. 35:53the same way on the chart sometimes if
  1076. 35:55you know how to read it.
  1077. 35:56>> So what you're basically describing here
  1078. 35:57is essentially what I do and I don't
  1079. 35:59want to make the conversation just about
  1080. 36:00my my beliefs, um, but that 's where I
  1081. 36:02found as a retail trader some edge and
  1082. 36:04when I speak to others, uh, that are
  1083. 36:06able to go into the wise like you are
  1084. 36:07with me today, it seems to be the common
  1085. 36:10answer is as a retail trader the edges
  1086. 36:12seem to lie on the lower time frames on
  1087. 36:13an intraday perspective or scalping.
  1088. 36:15But in the beginning you said day
  1089. 36:17trading is a is the wrong place to begin
  1090. 36:18or you don't like it.
  1091. 36:19>> I
  1092. 36:22I have seen more traders have more
  1093. 36:24success on slightly longer time frames.
  1094. 36:26Uh, and it it is interesting that you
  1095. 36:29know, we navigated around to basically
  1096. 36:31what you do. Like you know, you you
  1097. 36:33didn't tell me that. I didn't know that
  1098. 36:35it it just just by talking about where
  1099. 36:36the edges are, you have found it in
  1100. 36:38practice and you know, I found it in
  1101. 36:40practice and and theory too. So you you
  1102. 36:43are well aligned. But, uh
  1103. 36:46for me,
  1104. 36:49there was a point where I was I was
  1105. 36:50beyond obsessed with day trading. I
  1106. 36:52would trade the S&Ps. I would do a
  1107. 36:55little bit of music. I would And then I
  1108. 36:56would trade Europe at night. I would get
  1109. 36:58up and I don't know how I slept. Like,
  1110. 37:00you know, I add up the days the hours I
  1111. 37:02and it adds up to more than 24 hours. I
  1112. 37:04did that for a while.
  1113. 37:06And then I realized, you know,
  1114. 37:08it's not really how I want to be a
  1115. 37:09human. We're here. We don't know how
  1116. 37:11long. Like, uh plane could fall through
  1117. 37:14this right now while we're talking,
  1118. 37:15right? It could happen. Uh
  1119. 37:18you you never know the future. I don't
  1120. 37:20want to spend my entire life glued to
  1121. 37:21the screen. And if I'm going to day
  1122. 37:24trade well, and every now and then I get
  1123. 37:27I kind of get a thought, you know, maybe
  1124. 37:29I'd like to day trade again.
  1125. 37:30And my business partner, my wife are
  1126. 37:32like, okay. We we've we've seen this
  1127. 37:34before.
  1128. 37:35>> Mhm.
  1129. 37:35>> Go ahead. Go ahead. Come back and tell
  1130. 37:37me in a few months. And I did it. I'm
  1131. 37:39like, I this is not what I want to do. I
  1132. 37:40hate I I don't want to be glued to the
  1133. 37:42screen 24 hours a day. I don't have to.
  1134. 37:44I can make more money easier, certainly
  1135. 37:48for a unit of my time on a longer time
  1136. 37:50frame. But, I've also I think
  1137. 37:54the
  1138. 37:55everything's good and bad, right?
  1139. 37:56Everything's a double-edged sword. And
  1140. 37:58the advantage of day trading is you get
  1141. 38:01many times at bat. You have to make many
  1142. 38:03many decisions throughout the day. Uh
  1143. 38:05you know, a disciplined day trader on a
  1144. 38:08one-minute chart, you're probably making
  1145. 38:10what on a slow day three trades and
  1146. 38:12maybe up to two dozen on a busy day. Is
  1147. 38:14that roughly where you are?
  1148. 38:16>> So, just to briefly give you context, I
  1149. 38:18suppose it would help. Um I found the
  1150. 38:20sweet spot for me. I don't like look
  1151. 38:22because exactly what you're saying,
  1152. 38:23there is maybe an edge there, but it's
  1153. 38:24at the expense of your time and health
  1154. 38:26and and and everything else. Uh so, what
  1155. 38:29what I like when I started, I was
  1156. 38:30trading many asset classes. I realized I
  1157. 38:32want to focus on two.
  1158. 38:33>> Okay.
  1159. 38:33>> And I [clears throat] I wanted to trade
  1160. 38:34two things that are generally
  1161. 38:35correlated, but they do de-peg at times
  1162. 38:38because I can give meaning to that. So,
  1163. 38:39I ended up trading euro dollar, pound
  1164. 38:41dollar. And I've been doing that
  1165. 38:42probably for six years now exclusively.
  1166. 38:44I'll I'll one to four trades a week. Uh
  1167. 38:46busy periods will get maybe up to 10 a
  1168. 38:48week.
  1169. 38:49>> Okay.
  1170. 38:49>> Um, naturally if I'm in a trend, then I
  1171. 38:51can ride that trend. And what I found
  1172. 38:52the advantages of doing so, because I I
  1173. 38:54get what you're saying. Why don't I just
  1174. 38:56turn that read into a swing position?
  1175. 38:58But I found the benefit is if I can
  1176. 39:00sense price could go and make a lower
  1177. 39:01low in a bearish market, I can find the
  1178. 39:03heightened pockets of reaction, pockets
  1179. 39:06of probability.
  1180. 39:07>> Yeah.
  1181. 39:07>> And instead of me catching one big one
  1182. 39:08to five R, I can catch a few one to
  1183. 39:10threes along the way. So my trading
  1184. 39:12frequency is higher. My win rate is not
  1185. 39:13the best. It floats around 35 to 55.
  1186. 39:16>> Yeah, I don't care about your win rate
  1187. 39:17whatsoever.
  1188. 39:18>> my risk to reward can be one to threes
  1189. 39:20and my rate of return can be higher. And
  1190. 39:22I and I have a good trade frequency, I
  1191. 39:24think. And then I have my A plus setups
  1192. 39:26and my B plus and my C plus XYZ. So I
  1193. 39:28deploy them depending on recent
  1194. 39:30performance and I've architected it that
  1195. 39:32way.
  1196. 39:33Um,
  1197. 39:33>> I guess one of the challenges I would
  1198. 39:34throw out there is should you be doing
  1199. 39:35your C trade, your BC trades at all? Are
  1200. 39:38you better off increasing leverage on
  1201. 39:40your A trades? For many traders, that
  1202. 39:42turns out to be the answer.
  1203. 39:44>> yeah, big debate I have is discretionary
  1204. 39:45sizing. Do I size up on my high
  1205. 39:47conviction plays
  1206. 39:48or do I standardize? I've chosen to
  1207. 39:50standardize and I see the merits in
  1208. 39:51sizing up on on A plus setup.
  1209. 39:54But uh
  1210. 39:55we're talking about risk or exposure to
  1211. 39:56the market. So there's two ways to have
  1212. 39:57great exposure, sizing up on a high
  1213. 39:59conviction or a certain scenarios in the
  1214. 40:01market taking my C plus setup, which has
  1215. 40:03a lower win rate. But if I know it
  1216. 40:06happened to me in September actually,
  1217. 40:07I went through a hot period, not through
  1218. 40:09skill, it's just the distribution of
  1219. 40:11outcomes, a 50% win rate system can have
  1220. 40:14a recent performance of 80% win because
  1221. 40:16it just the wins align and now the
  1222. 40:17losses may may follow. So as a result of
  1223. 40:20that, if I have my
  1224. 40:22trajectory of my expectancy and it's and
  1225. 40:24it's just up into the top right,
  1226. 40:25hopefully. And then I have a period
  1227. 40:27where it overshoots. Now it doesn't mean
  1228. 40:29I've beaten my edge and now I'm going to
  1229. 40:31have a new like superstar edge. It's
  1230. 40:33just I've overperformed. But I realize
  1231. 40:35when the when the market is hot or I'm
  1232. 40:36hot at the moment, I can now add in a
  1233. 40:39little bit more risk through riskier
  1234. 40:41setups, which I wouldn't ordinarily take
  1235. 40:43because my relative recent win rate is
  1236. 40:4580.
  1237. 40:46It should be a 40. So now I can afford
  1238. 40:48to bring it down gradually, and I know
  1239. 40:50I'm calibrating it. So when I start to
  1240. 40:51hit the average of 50 in my last 15
  1241. 40:54trades, get back to my A+ setups. So the
  1242. 40:57as I revert back to my mean of my
  1243. 40:58trajectory of my expectancy,
  1244. 41:00uh I can control when to deploy these
  1245. 41:02higher risk trades.
  1246. 41:04Uh instead of sizing up now, I can
  1247. 41:05actually get uh greater rate of return
  1248. 41:07through greater trade frequency
  1249. 41:09temporarily. That's how I've navigated
  1250. 41:10it. I'd love to hear your thoughts
  1251. 41:12actually. Would I be better off just
  1252. 41:13sizing up on high conviction and having
  1253. 41:15a lower general trade frequency?
  1254. 41:17>> So
  1255. 41:19if on the low side you're doing four
  1256. 41:22trades a week,
  1257. 41:24you're you're kind of
  1258. 41:27tilted toward that longer holding period
  1259. 41:29that I was talking about I think, right?
  1260. 41:31Or are are these are these four short
  1261. 41:33>> I'm trading FX. So it doesn't have the
  1262. 41:35baked in bias, so it doesn't necessarily
  1263. 41:36serve me to just
  1264. 41:38turn it into a long-term hold. So I
  1265. 41:39rarely hold a trade overnight.
  1266. 41:41>> Okay.
  1267. 41:41>> It would be usually a session reaction
  1268. 41:43for my TP1, the average move of a
  1269. 41:45session, and then my usually my TP2 and
  1270. 41:47final TP is average movement of the day.
  1271. 41:49I might leave a runner if I see that
  1272. 41:51there's a long-term trend in play, but
  1273. 41:53generally not.
  1274. 41:54>> I generally find that most traders, you
  1275. 41:57know, this ABC trade book is a common
  1276. 42:00way to think about it. Most traders are
  1277. 42:01better off not doing their B and C
  1278. 42:02trades. They're
  1279. 42:04But what you're describing is exactly If
  1280. 42:07you If your C trades are trades that
  1281. 42:09only happen are more likely to happen in
  1282. 42:12certain market environments that you
  1283. 42:14reach these environments
  1284. 42:15>> That and I define it by lower win rates.
  1285. 42:17Like it it has a 1 to 5 risk reward
  1286. 42:19potential, um but it it has a 40% win
  1287. 42:22rate. I don't want to start with that.
  1288. 42:24>> strike me as a C trade, you know, the
  1289. 42:25the the stats
  1290. 42:26>> It still has an edge, but it's just not
  1291. 42:27as good as the other ones that might
  1292. 42:28appear twice a month where it can give
  1293. 42:30me a 1 to 7 and it can give me a 80 % or
  1294. 42:32not 80 probably around 65% win rate.
  1295. 42:34These are the ones I would love to only
  1296. 42:36focus on, but if I exclusively trade
  1297. 42:37them, my rate of return suffers because
  1298. 42:39I am taking a few trades a month. And if
  1299. 42:42I have a bad month, it just
  1300. 42:44to my personality, I don't like being
  1301. 42:45wrong.
  1302. 42:45>> Sure. Sure. No, I understand that. Um
  1303. 42:49feels to me without like really getting
  1304. 42:51into your numbers like you've probably
  1305. 42:53laid this out in a way that makes sense
  1306. 42:54for you.
  1307. 42:55I would be I would be
  1308. 42:57very hesitant for you to come away from
  1309. 42:59this interview and go make changes based
  1310. 43:00to based on anything that I'm saying
  1311. 43:02because you are you've already asked a
  1312. 43:04lot of the questions that I would ask
  1313. 43:06and you seem to have good answers to
  1314. 43:07them and I think it sounds like you
  1315. 43:09found what's working for you right now.
  1316. 43:12>> The the the interview has kind of
  1317. 43:13flipped a little
  1318. 43:14Let's uh let's put it back into towards
  1319. 43:16yourself. What would your equivalent be
  1320. 43:17of what I've just described of
  1321. 43:19your KPIs, the way you deployed risk,
  1322. 43:21how you built an edge around your
  1323. 43:23temperament, which it seems like a a
  1324. 43:24journey you had after your day trading
  1325. 43:26uh era. What does life look like now in
  1326. 43:28the markets?
  1327. 43:29>> my
  1328. 43:31you know I'm a I'm a person who
  1329. 43:35really feeds off of obsession. If I do
  1330. 43:37something
  1331. 43:39I I I have two speeds, either full stop
  1332. 43:42or full speed ahead.
  1333. 43:43And what I have found with the markets
  1334. 43:46is
  1335. 43:47while I have been involved in trading in
  1336. 43:50the markets, I've gone through periods I
  1337. 43:52haven't been able to maintain for my
  1338. 43:54entire career full focus. I think if I
  1339. 43:57had it would have been very unhealthy.
  1340. 43:58And I have recently been in a period of
  1341. 44:02the major life transition where I
  1342. 44:04haven't been trading quite as much. I
  1343. 44:06haven't been taking as much risk. I've
  1344. 44:07been uh doing more
  1345. 44:11more simple trades, more simple only
  1346. 44:14doing the trades that really jump out
  1347. 44:16and say you know this is a very clear
  1348. 44:18pattern that you have to take. And also
  1349. 44:20to be one thing we maybe should discuss
  1350. 44:22is I don't think the visual prettiness
  1351. 44:25of a chart picture is any
  1352. 44:28judge of its outcome. You know so I used
  1353. 44:29to think that. Some people think that.
  1354. 44:31But if you
  1355. 44:33do a few thousand trades and separate
  1356. 44:35them into
  1357. 44:36patterns that looked really good and
  1358. 44:37ones that don't and again do they does
  1359. 44:39the outcome look different and you know
  1360. 44:40you can do tests like that when you have
  1361. 44:42data.
  1362. 44:43Um
  1363. 44:44now my
  1364. 44:46I think I'm looking at a future because
  1365. 44:48I 'm also thinking you know recently I
  1366. 44:50had my 50th birthday a few years ago and
  1367. 44:52I 'm thinking about how am I going to
  1368. 44:54trade in the future.
  1369. 44:55And it's not going to look like how I
  1370. 44:57traded in my twenties and probably not
  1371. 44:59going to look like how I traded 10 years
  1372. 45:02ago.
  1373. 45:02So my desire is to move into more of a
  1374. 45:08I've always been a discretionary trader
  1375. 45:10that has a quantitative foundation.
  1376. 45:13You know, that I I've asked these kinds
  1377. 45:14of questions, the things that you ask
  1378. 45:16about Fibonacci, things I've had you
  1379. 45:18know, every pattern I trade, if I see a
  1380. 45:21market that makes a big move and then
  1381. 45:23makes a consolidation, is there an edge
  1382. 45:25to buying at the bottom of that
  1383. 45:27consolidation? How can I define that uh
  1384. 45:29if I have and some surprising things
  1385. 45:32work. If I have a trend line, I I draw a
  1386. 45:34trend line and then I draw a parallel
  1387. 45:36trend line, is there an edge to
  1388. 45:38executing against that parallel trend
  1389. 45:40line?
  1390. 45:40There is.
  1391. 45:41>> Okay.
  1392. 45:41>> Why? I don't know. Volatility probably
  1393. 45:44has something to do with random walk
  1394. 45:45volatility, but there is and then the
  1395. 45:49you know, maybe the higher time frame
  1396. 45:50influence of the I I I I do know it's
  1397. 45:52there. It's but it's a you know, issue
  1398. 45:55of shorter time frame mean reversion
  1399. 45:57playing into longer time frame momentum.
  1400. 45:59Uh
  1401. 46:00so that is how I have traded, but I
  1402. 46:03think I'm looking at doing something
  1403. 46:04like more of a weighted alpha kind of
  1404. 46:07approach where and this is actually a
  1405. 46:09loop that we should close. Your question
  1406. 46:10about the 200-day moving average. Um
  1407. 46:14I can
  1408. 46:16for me to have something that I can
  1409. 46:17trade as a standalone, it has to be
  1410. 46:19quite a good edge, right? Uh it has to
  1411. 46:21have a certain expectancy, has to occur
  1412. 46:25often. You know, if if it's a great
  1413. 46:26trade that happens once a year, I might
  1414. 46:28have three bad years trading it, right?
  1415. 46:30You you just don't So you have to have
  1416. 46:31you have all these things that play in
  1417. 46:33to the system that's going to be the
  1418. 46:34system you trade.
  1419. 46:35One of the things you can do as a
  1420. 46:37quantitative trader though and this
  1421. 46:39for discretionary for retail traders
  1422. 46:41this I think this is going to be
  1423. 46:42important because people don't on the
  1424. 46:44retail side typically think about this.
  1425. 46:46I can have
  1426. 46:48a handful of
  1427. 46:50systems edges that kind of suck,
  1428. 46:53but they have a small edge. There's no
  1429. 46:55way I could trade
  1430. 46:56And, you know, I might have
  1431. 47:00It's like this. Uh,
  1432. 47:01let's say you have 50 friends who
  1433. 47:05are kind of good at guessing the
  1434. 47:06direction of the market, but none of
  1435. 47:08them like would really make that much
  1436. 47:09money. You call them up every morning
  1437. 47:12and you say, is the market going to go
  1438. 47:13up or down today? And then you average
  1439. 47:15them.
  1440. 47:17And
  1441. 47:19on days where most of them are telling
  1442. 47:21you the market's going up, you would
  1443. 47:22probably find you had a much stronger
  1444. 47:25signal than you would from any one of
  1445. 47:27those people. If you do it, any one of
  1446. 47:28those people tells you you're not going
  1447. 47:30to make money. But you average them
  1448. 47:32together. This is what happens in a lot
  1449. 47:35of quant trading. You take these alphas,
  1450. 47:37and they have to be legitimate alphas,
  1451. 47:39and you said an important word earlier,
  1452. 47:41they have to be uncorrelated. And the
  1453. 47:43problem is not that many things are
  1454. 47:45truly uncorrelated because let's say I
  1455. 47:48want to trade mean reversion. Great.
  1456. 47:49Okay. So
  1457. 47:52a signal might be a certain RSI level,
  1458. 47:55might be a momentum divergence, might be
  1459. 47:57a big standard deviation move down, it
  1460. 47:59might be crossing a Bollinger band, it
  1461. 48:01might be breaking under a five or 10 or
  1462. 48:0320 day high. All of these would be ways
  1463. 48:05that we would measure a big move down,
  1464. 48:07right? They're all highly correlated. So
  1465. 48:10so if I if I think that I I just told
  1466. 48:13you six different ways that I can put in
  1467. 48:14my average alpha, no, I it's it's like
  1468. 48:17me calling the same person six times in
  1469. 48:19the average. So you have to understand,
  1470. 48:23you have to find these edges, and you
  1471. 48:26have to they have to be uncorrelated,
  1472. 48:29you average them together, and you get a
  1473. 48:30much stronger signal than you would from
  1474. 48:31any of them. None of them
  1475. 48:33individually have to be
  1476. 48:36standout.
  1477. 48:38>> I don't know why, but that bothers me
  1478. 48:39from a perspective of
  1479. 48:41why does that work?
  1480. 48:42>> Yep. Well, it's uh
  1481. 48:46I think it has to do with the wisdom of
  1482. 48:47crowds. I don't know if you know that
  1483. 48:49book, but like one of the
  1484. 48:51since we're sitting here in the middle
  1485. 48:52of farmland in America, uh
  1486. 48:56we have these county fairs. And one of
  1487. 48:58the things that happens at these fairs,
  1488. 49:01they would have an animal. They would
  1489. 49:02have, you know, a bull or a pig. And
  1490. 49:05people would
  1491. 49:06put
  1492. 49:07they they would bet on the weight of the
  1493. 49:09animal. If they get the weight right,
  1494. 49:12they get the animal. And what you find
  1495. 49:15is that if you average up all of those
  1496. 49:18guesses, they're actually quite close in
  1497. 49:21most cases to what the animal actually
  1498. 49:22weighs. So what's actually happening
  1499. 49:24there?
  1500. 49:26You have people like me and you know, I
  1501. 49:27assume maybe you who have no idea what
  1502. 49:29they are. So like we're just writing
  1503. 49:30numbers down. My guess is not good. Your
  1504. 49:32guess is not good. But
  1505. 49:35people like us who don't know anything,
  1506. 49:36those guesses are going to average out
  1507. 49:38to kind of zero. And then you have the
  1508. 49:40handful of people who really know what
  1509. 49:41they're looking at, who have intelligent
  1510. 49:43guesses and those actually. So it's kind
  1511. 49:46of like that with the Alphas, you know,
  1512. 49:47the ones that if they're weak working
  1513. 49:51together,
  1514. 49:52the ones that
  1515. 49:54they'll tend to average each other out
  1516. 49:56until several of them point in the same
  1517. 49:57direction.
  1518. 49:58So you can't do this on a discretionary
  1519. 50:00basis to be clear.
  1520. 50:02>> Oh, my question was actually going to be
  1521. 50:05a retail trader listening it might be
  1522. 50:07like, okay, I I I like Fibonacci, but
  1523. 50:09it's not good enough. So let me cross
  1524. 50:10that over with the trend line and an EMA
  1525. 50:12and a doji candle and a head and
  1526. 50:13shoulder pattern. All of a sudden I've
  1527. 50:15got a cocktail of confluences.
  1528. 50:17>> Yeah.
  1529. 50:17>> None of them were necessarily causative
  1530. 50:18or had an edge, but in combination
  1531. 50:21it might be.
  1532. 50:22>> I I don't I've never seen that work. I
  1533. 50:26because first of all, many of those
  1534. 50:27things are correlated, right?
  1535. 50:29The Fibonacci level and the trend line
  1536. 50:31and the EMA and you know, throw a band
  1537. 50:34in there. You've you've moved to a level
  1538. 50:37and those things are probably and are
  1539. 50:39you know, the argument of course is
  1540. 50:40well, if the trend line's at the level,
  1541. 50:41maybe it's more meaningful and you know,
  1542. 50:43we we can test that and look at it.
  1543. 50:46I
  1544. 50:48in the in the discretionary world
  1545. 50:51this is there's also this there's this
  1546. 50:53argument against about should we have
  1547. 50:56confluence and or confirmation. Those
  1548. 50:59are kind of two separate things, right?
  1549. 51:01Do you want to buy in anticipation of
  1550. 51:03the move or sell in anticipation or do
  1551. 51:05you want to wait for the move wait for
  1552. 51:07the train to start moving and then you
  1553. 51:08jump on board. Both of those things work
  1554. 51:12but with the with the confluence
  1555. 51:16people are many times just measuring the
  1556. 51:18same things and they're using things
  1557. 51:19that don't work and is it possible is it
  1558. 51:22possible that something does not work at
  1559. 51:24all and when you put it with something
  1560. 51:27else that then it magically works.
  1561. 51:32Probably not. It it might it might be
  1562. 51:35possible that we have a signal that has
  1563. 51:37no edge and then we put something under
  1564. 51:39it that filters out a bunch of the
  1565. 51:41signals and we have a
  1566. 51:42I haven't really seen that work very
  1567. 51:45reliably. It's theoretically possible,
  1568. 51:46but the key to making this averaged
  1569. 51:49thing work is that all of the each of
  1570. 51:52the pieces has to have an edge
  1571. 51:54separately and they need to be
  1572. 51:56uncorrelated. Most of the things like
  1573. 51:58you know the Fibonacci levels that we're
  1574. 52:00talking about
  1575. 52:01there's not an edge there in the first
  1576. 52:02place.
  1577. 52:03>> you remind me why uncorrelated was
  1578. 52:04necessary because if I'm stacking this
  1579. 52:06EMA, trend line, Fibonacci, da da da,
  1580. 52:08yes they're all correlated telling me
  1581. 52:10the same thing, but that's the point.
  1582. 52:11They are
  1583. 52:12adding additional
  1584. 52:13>> for the you're looking for the
  1585. 52:14correlation and you're assuming the
  1586. 52:15correlation.
  1587. 52:16>> Assuming they are adding odds stacking
  1588. 52:17odds in my favor.
  1589. 52:18>> You're just measuring the same thing a
  1590. 52:20different way.
  1591. 52:21>> Uh-huh.
  1592. 52:21>> You're you're
  1593. 52:23I that that's the argument against it
  1594. 52:25you because everything we look at in the
  1595. 52:27market is you know if if if you really
  1596. 52:31want to blow your mind
  1597. 52:32going back to first principles which is
  1598. 52:34a phrase you said everything we look at
  1599. 52:36in the market is something that's
  1600. 52:39happening in the market, but it's also
  1601. 52:41the structure that we're imposing on the
  1602. 52:43market. It's the way that we're
  1603. 52:44measuring it. Even just a chart you can
  1604. 52:46say no there's no structure Adam. I'm
  1605. 52:48just drawing a line of a like every
  1606. 52:50trade you know a one tick line chart.
  1607. 52:52That's still a structure, right? Because
  1608. 52:54you're still you still have an axis and
  1609. 52:56certainly if we put it into bars or
  1610. 52:58candles
  1611. 52:59we've put a structure around it and then
  1612. 53:01trend
  1613. 53:02you can see everything else right how we
  1614. 53:04measured that. Uh, RSI, trend line,
  1615. 53:07Fibonacci, you're putting structure on
  1616. 53:09it. So you're you're just measuring it
  1617. 53:11and is there that fundamental you know,
  1618. 53:14the market made this move down and then
  1619. 53:16all of these things are going to speak.
  1620. 53:20The you know, the the quantitative thing
  1621. 53:22is see if you have a regression model uh
  1622. 53:25you know, people who have more math than
  1623. 53:27I do understand this intuitively. And if
  1624. 53:29those inputs into the regression model
  1625. 53:30are highly correlated, they make your
  1626. 53:32model look a lot better than it is. They
  1627. 53:34make your confidence in the model, but
  1628. 53:36it's not actually that good because your
  1629. 53:37inputs, it's one of the things you're
  1630. 53:38looking at specifically and most models
  1631. 53:41is you want to see those inputs are not
  1632. 53:42correlated. And in reality, everything's
  1633. 53:44kind of correlated, so you don't you
  1634. 53:46never get to zero.
  1635. 53:48>> Yeah, I guess to visualize it, it's like
  1636. 53:50all these confluences were like Venn
  1637. 53:51diagrams, but you're just adding more
  1638. 53:53circles, but the crossover period is
  1639. 53:55what's relevant. And as opposed to rungs
  1640. 53:58on a ladder, they're not climbing you
  1641. 53:59higher. It's more like the Venn diagram.
  1642. 54:01Okay. I'm I'm also not curious on
  1643. 54:03invalidations because we've spoken about
  1644. 54:05reasons price could do something
  1645. 54:07>> Yeah.
  1646. 54:08>> and confluences and then uh if they're
  1647. 54:10too correlated X, Y, Z,
  1648. 54:12you could you could do this process, but
  1649. 54:14then I guess what would be more relevant
  1650. 54:15is two things that appear optically
  1651. 54:17similar or parameter similar, but you
  1652. 54:20would take one and you wouldn't take the
  1653. 54:21other. What would certain invalidations
  1654. 54:22look like?
  1655. 54:24>> For me, um
  1656. 54:29so understand and we keep talking about
  1657. 54:30buying, maybe we should talk about
  1658. 54:31selling to make it symmetrical, but I'll
  1659. 54:33keep it on buying. If if I'm looking to
  1660. 54:35buy a consolidation,
  1661. 54:37I don't want to see that it makes many
  1662. 54:40probes to the upside and is rejected.
  1663. 54:44So visually, if you're looking at a
  1664. 54:46chart,
  1665. 54:47I don't want that upper boundary to be
  1666. 54:49really ragged. The market makes a swing
  1667. 54:51high and then comes down. I don't like
  1668. 54:54if you're going above that swing high
  1669. 54:56and getting rejected there because
  1670. 54:57there's price we're back to Wyckoff,
  1671. 54:58right? Uh, on the other hand, I'm really
  1672. 55:01happy to see that ragged bottom. You
  1673. 55:03know, I'd like to see the the contained
  1674. 55:05and then it tries to sell off and they
  1675. 55:07buy it right back up. So that kind of
  1676. 55:10asymmetry. It's fine if the bottom
  1677. 55:12smooths too, but I certainly uh if I've
  1678. 55:15been watching for an entry price and a
  1679. 55:18pullback and it breaks out and then
  1680. 55:21fails, in most cases I'm no longer
  1681. 55:23interested. I you know I I have a very
  1682. 55:26easy time killing setups. I I have a you
  1683. 55:30know I can be really excited I don't get
  1684. 55:32excited about anything, but I can be
  1685. 55:32really excited about a setup and then it
  1686. 55:34does something I don't even want to talk
  1687. 55:35It's dead to me. I'm done with it. Um
  1688. 55:38the
  1689. 55:39sometimes it just goes on too long. You
  1690. 55:42know there should be my my thinking and
  1691. 55:44here we're back to stories. My thinking
  1692. 55:46is that
  1693. 55:48if I have identified one of these places
  1694. 55:50there's going to be good continuation
  1695. 55:51there should be some urgency behind it.
  1696. 55:53I I shouldn't see you know if there is
  1697. 55:55if if we're talking about a commodity
  1698. 55:57and there's a supply shortage, I should
  1699. 55:59not see it make a big move up
  1700. 56:01and then just kind of sit sideways for
  1701. 56:02months. That doesn't really say there's
  1702. 56:04any kind of urgency there. Same thing if
  1703. 56:06it's a one minute chart, you know we
  1704. 56:08just just scale it differently. I want
  1705. 56:10to see that there's urgency.
  1706. 56:11>> Does that imply that you got in for a
  1707. 56:13reason, whatever that reason may be, and
  1708. 56:16volume or momentum is necessary and
  1709. 56:18therefore a sense of invalidation is
  1710. 56:19lack of participation in your direction.
  1711. 56:21Would you therefore
  1712. 56:23act upon that in the sense of you're
  1713. 56:24already in a position, you have your
  1714. 56:25parameters of take profit entry and stop
  1715. 56:27loss, price is just around your entry,
  1716. 56:29it's a little bit in profit. Will you
  1717. 56:30say okay upon a kind of time decay this
  1718. 56:33is taking too long long. It could still
  1719. 56:35play out. It's just not happening in the
  1720. 56:37time frames I needed to, but it hasn't
  1721. 56:39invalidated. It hasn't hit my stop. The
  1722. 56:41idea is still there. Would you exit that
  1723. 56:43position?
  1724. 56:43>> Yeah. Yeah, I I would. That that is one
  1725. 56:46of the
  1726. 56:49So you know when I told you that I
  1727. 56:50basically traded the same way for
  1728. 56:52decades
  1729. 56:53trading the basic structure, these kinds
  1730. 56:56of questions exactly how do you manage a
  1731. 56:58trade? How do you get out when it's
  1732. 56:59working? How do you get out when it's
  1733. 57:01not? How do you know when it's not
  1734. 57:02working?
  1735. 57:03What do you do if it's not working and
  1736. 57:04then it's working again you know how do
  1737. 57:05you make These are the kinds of
  1738. 57:06decisions that
  1739. 57:09do evolve and change and there would
  1740. 57:12have been a time uh
  1741. 57:14when I was a day trader, the answer to
  1742. 57:16that question was automatically
  1743. 57:18instantly I'm out. I'm looking for any
  1744. 57:21excuse as a day trader to say the trade
  1745. 57:23is not working, I am done. And you know
  1746. 57:27you you learn this by getting kicked in
  1747. 57:28the head doing doing it the wrong way
  1748. 57:30enough times. As soon as you have a
  1749. 57:32little bit of a a little bit of hope you
  1750. 57:34know a little bit let's just give it a
  1751. 57:35little more time. Sometimes it'll work
  1752. 57:37and you'll remember those times but if
  1753. 57:39you run your stats far more often you
  1754. 57:41need to get out.
  1755. 57:43But on on the daily time frame it has
  1756. 57:46paid to be a little bit stickier and I
  1757. 57:48think this is one of the ways the market
  1758. 57:49has changed in the past
  1759. 57:51>> three years. It's no longer pain to be
  1760. 57:53sticky on the daily time frame if
  1761. 57:55something doesn't look good.
  1762. 57:57Better situation.
  1763. 57:58>> I was I was thinking when to bring up
  1764. 57:59this topic but I feel like you've
  1765. 58:02there's many roads in this conversation
  1766. 58:03that have led to this question.
  1767. 58:05You were talking about the trend line
  1768. 58:06and having a trend line parallel to it
  1769. 58:08and there seems to be something there. I
  1770. 58:09took note of it. You were mentioning how
  1771. 58:11when you get to the range high and
  1772. 58:13you're getting rejection as opposed to
  1773. 58:14movement that would be some sort of
  1774. 58:15invalidation. I took note of that cuz
  1775. 58:17what would that mean? And then when
  1776. 58:18you're talking about okay you got into
  1777. 58:20an you got in for a certain reason but
  1778. 58:21it's not moving as quick as you would
  1779. 58:22need to
  1780. 58:24and you would you would take off you
  1781. 58:25take off the position even though
  1782. 58:26nothing has changed apart from time has
  1783. 58:28elapsed. I took note of that.
  1784. 58:30These things all
  1785. 58:31bring me to
  1786. 58:33I don't know the term for it but it
  1787. 58:34might be a sense of market impact
  1788. 58:37kind of a reflection of order flow where
  1789. 58:39there are moments in the market that are
  1790. 58:40highly reactionary I found. Let's say
  1791. 58:43just to draw a visual
  1792. 58:44there there's grades to it that I have
  1793. 58:46but a high of previous day
  1794. 58:47we could call that some sort of
  1795. 58:49resistance. I've found if price sweeps
  1796. 58:51it pushes higher
  1797. 58:53naturally you invite in breakout traders
  1798. 58:54they're going to look to buy that and
  1799. 58:56then also you take the stops of the
  1800. 58:57people that were holding their cells in
  1801. 58:59that position. So this becomes a high
  1802. 59:00impact or high inflection point. And
  1803. 59:02I've noted that when you sweep it you
  1804. 59:05get a reaction down that might not be
  1805. 59:07for the rest of the day rest of the rest
  1806. 59:08of the month but for a duration it will.
  1807. 59:11>> Yep.
  1808. 59:11>> And that's basically the center of my
  1809. 59:13edge. Now I might be not it's not just
  1810. 59:14high previous day, it could be
  1811. 59:16the high of previous week, it could also
  1812. 59:17be a structural point. There's many
  1813. 59:18levels to it, but that's basically an
  1814. 59:20extension of my edge. And then I've
  1815. 59:22noticed like the bigger zones like a
  1816. 59:24daily high uh sorry
  1817. 59:26let's say a daily impulse, a lower high
  1818. 59:27to lower low. At the extreme of it, if
  1819. 59:29we take that high, it gives a bit of a
  1820. 59:31bigger reaction as opposed to just in
  1821. 59:33New York session you take London's high,
  1822. 59:34which gives a smaller reaction.
  1823. 59:35>> Mhm.
  1824. 59:36>> What is going on here? What what am I
  1825. 59:38describing?
  1826. 59:38>> I think you've like you laid out the
  1827. 59:40dynamics perfectly. That's exactly what
  1828. 59:42happens.
  1829. 59:43>> I mean more of like why is it happening?
  1830. 59:45>> Because you are the market is always
  1831. 59:48probing into areas to see if there's
  1832. 59:51real interest. And I couldn't have said
  1833. 59:54it better than you did. In my first
  1834. 59:56book, I laid out six basic price
  1835. 1:00:00patterns, what you're describing, your
  1836. 1:00:02edge was the first one that I put in the
  1837. 1:00:03book. This idea of penetrating a level
  1838. 1:00:06and then failing. So you know one of the
  1839. 1:00:10things that has happened, markets always
  1840. 1:00:12markets are looking for liquidity.
  1841. 1:00:14Markets want to make trades. They the
  1842. 1:00:17exchange exists to facilitate trade, to
  1843. 1:00:20make trades happen. So markets are
  1844. 1:00:22always looking for places to make
  1845. 1:00:24trades. And the market is if we think of
  1846. 1:00:27the market as a conscious being, the
  1847. 1:00:29market's going to probe above those
  1848. 1:00:31highs and to say is there really buying
  1849. 1:00:33pressure here? Certainly there are there
  1850. 1:00:35are orders, you know I can I know I can
  1851. 1:00:36go get those orders executed because
  1852. 1:00:38people have shorts, they have stops, and
  1853. 1:00:40there are the hopeful breakout traders.
  1854. 1:00:42But how much buying pressure is how do I
  1855. 1:00:44know? Well let's go execute there and
  1856. 1:00:46see. And if there wasn't much, then
  1857. 1:00:48those that those people are going to
  1858. 1:00:50sell uh you know the the hopeful
  1859. 1:00:52breakout traders when their hope is
  1860. 1:00:54gone, they sell and then you kind of get
  1861. 1:00:56this cascade of selling. For me, there's
  1862. 1:00:58also a reaction after that. So you have
  1863. 1:01:01that penetration versus if you have
  1864. 1:01:03acceptance where you where you go up
  1865. 1:01:05above the high and turns out oh there's
  1866. 1:01:08actually real buying pressure here,
  1867. 1:01:10there is continuation. If you have that
  1868. 1:01:12failure scenario,
  1869. 1:01:14then there's
  1870. 1:01:15potentially kind of a second move,
  1871. 1:01:17right? You have you have the failure,
  1872. 1:01:19the market goes back below the high.
  1873. 1:01:21Everybody's like, oh, what happened,
  1874. 1:01:22right? And you sit there for a minute
  1875. 1:01:24and then if more sellers come in, you
  1876. 1:01:26may get another leg down. That's also
  1877. 1:01:28quite uh reliable tradable move
  1878. 1:01:30sometimes. And then of course, you know,
  1879. 1:01:32it's uh
  1880. 1:01:34back to your question about me what I'm
  1881. 1:01:37doing and when I'm in a trade, when I'm
  1882. 1:01:39thinking about a trade, every tick that
  1883. 1:01:41comes in is data.
  1884. 1:01:43And
  1885. 1:01:44I'm analyzing every piece of data that
  1886. 1:01:46comes in. I'm probably analyzing most of
  1887. 1:01:48it wrong, but I'm getting enough right
  1888. 1:01:50to be able to be on the right side of
  1889. 1:01:52the market more often than not. And so,
  1890. 1:01:56you know, the market comes in and now
  1891. 1:01:58it's processing that move and now if the
  1892. 1:02:01market sits near that high, you know,
  1893. 1:02:02now maybe we're sitting you know, you
  1894. 1:02:04can establish a short there against that
  1895. 1:02:06high, but the more time you sit there,
  1896. 1:02:09the probabilities are not really
  1897. 1:02:10favoring that drop. They're favoring
  1898. 1:02:12another trip to this high.
  1899. 1:02:13>> I like what you mentioned before of
  1900. 1:02:14confluence versus confirmation. And I
  1901. 1:02:17I've always toyed with which one is
  1902. 1:02:19better. Around these scenarios, you have
  1903. 1:02:21a liquidity point. Price shoots above
  1904. 1:02:23it. Naturally, this is a dangerous place
  1905. 1:02:25to be because it is by definition an
  1906. 1:02:26inflection point. You could get buyers
  1907. 1:02:28stepping in, it goes up. It could also
  1908. 1:02:29be a rejection point, it goes down. How
  1909. 1:02:31do we know which one? So, there's some
  1910. 1:02:32things you could do to I guess put put
  1911. 1:02:34the odds in your favor, but I I found
  1912. 1:02:36it's a bit more dangerous to enter
  1913. 1:02:37there, but when I see the rejection and
  1914. 1:02:39the sellers are taking strength, whether
  1915. 1:02:40it's defined by a M1 break of structure
  1916. 1:02:42or XYZ parameters, then upon certain
  1917. 1:02:45parameters, I'll then enter the short.
  1918. 1:02:47But the issue is it's hard to know if
  1919. 1:02:48this is is it going to be a sustained
  1920. 1:02:50move, which is why I can't turn that
  1921. 1:02:51into a swing trade even though I'm in a
  1922. 1:02:53bearish market and I found a sell at the
  1923. 1:02:55top. I don't know if I can hold it
  1924. 1:02:57because it could chop around and
  1925. 1:02:58invalidate the next day.
  1926. 1:02:59Uh so, what I've kind of found a
  1927. 1:03:01parameter for me that helps me is I just
  1928. 1:03:03break even super fast. I would just like
  1929. 1:03:05literally as soon as I can. It might tap
  1930. 1:03:07me out here and there. It does happen a
  1931. 1:03:08lot and that is frustrating, but a lot
  1932. 1:03:10of the times I can just take a shot not
  1933. 1:03:13knowing really what's going on, but it
  1934. 1:03:15just if I'm protected, I'm risk free and
  1935. 1:03:16then it happens to play. And uh if I
  1936. 1:03:18look back in my last year, around 30% of
  1937. 1:03:21my profit was trades where I got the
  1938. 1:03:23direction wrong. I was selling and the
  1939. 1:03:25the week ended up bullish, the day ended
  1940. 1:03:26up bullish, but I caught a reaction,
  1941. 1:03:28broke even fast, and let it run to
  1942. 1:03:30targets. So I was wrong, but yet
  1943. 1:03:32captured a profit. Seems weird, but that
  1944. 1:03:34was 30% of my profit for the last year.
  1945. 1:03:36>> that those are my favorite trades to be
  1946. 1:03:37to be wrong and make money. Yeah.
  1947. 1:03:39>> Yes.
  1948. 1:03:40Um I I'm also curious here because yes,
  1949. 1:03:42I'm I'm at a younger or a different
  1950. 1:03:44stage in my life compared to yourself,
  1951. 1:03:45but I also find that finding alpha
  1952. 1:03:47enough or an edge enough is so
  1953. 1:03:49difficult.
  1954. 1:03:50Most people are not in a place of luxury
  1955. 1:03:52or abundance to say, let me pick the
  1956. 1:03:53edge that suits me the most. How have
  1957. 1:03:55you found yourself in a position where
  1958. 1:03:57I don't know how to turn my approach to
  1959. 1:03:59a swing approach and and take it a bit
  1960. 1:04:01more easy. How have you been able to
  1961. 1:04:03reshape an edge that would be
  1962. 1:04:05draining on your time and consuming you
  1963. 1:04:06and your let's say obsessive personality
  1964. 1:04:08that you have to now be like, I'm
  1965. 1:04:09getting older now and I want to be less
  1966. 1:04:11stressed, so I'm going to now turn it
  1967. 1:04:13into a longer time horizon hold. It
  1968. 1:04:15feels like you're just going to be
  1969. 1:04:16looking for a whole new edge and a whole
  1970. 1:04:17new system.
  1971. 1:04:18>> I don't I don't think
  1972. 1:04:21you get to do what you're suggesting. I
  1973. 1:04:23I don't think you get to take your
  1974. 1:04:25specific edge and make it make it a
  1975. 1:04:27longer term hold. So, you know, back to
  1976. 1:04:30kind of themes we run through this
  1977. 1:04:32interview, we're talking about mean
  1978. 1:04:34reversion and momentum. And there are
  1979. 1:04:35traders who trade, broadly speaking,
  1980. 1:04:38mean reversion or in your a mean
  1981. 1:04:39reversion trader.
  1982. 1:04:40In general, and when you say in general,
  1983. 1:04:43you're going to make a lot of bad
  1984. 1:04:44generalizations, but in general, mean
  1985. 1:04:46reversion traders need to take profits
  1986. 1:04:48very quickly.
  1987. 1:04:49As soon as there is money on the table,
  1988. 1:04:51you put it in your pocket and you're
  1989. 1:04:53done. Momentum traders,
  1990. 1:04:55generally speaking, can be a little bit
  1991. 1:04:57stickier, you know, like I'll I'll let
  1992. 1:04:59it ride. I give I give it a little bit
  1993. 1:05:00more time. Um
  1994. 1:05:03you don't get to decide you want to play
  1995. 1:05:07for a bigger move because the another
  1996. 1:05:10way to think about it is every signal
  1997. 1:05:13that we would have and to talk broadly
  1998. 1:05:14in signals,
  1999. 1:05:16I think, and this may not be accurate,
  2000. 1:05:18but I think they have an
  2001. 1:05:21a limited influence over the market.
  2002. 1:05:24Both in terms of time and magnitude. So
  2003. 1:05:26I do not believe yes, we can go back 20
  2004. 1:05:29years and we can find the intraday low
  2005. 1:05:32from the British pound that held for
  2006. 1:05:35years and you can say that was the
  2007. 1:05:37moment that pattern help
  2008. 1:05:39No, it didn't. You know, I I I I don't
  2009. 1:05:41believe that and that's just I guess
  2010. 1:05:43it's a belief. But your
  2011. 1:05:46mean reverting patterns have a
  2012. 1:05:48relatively short duration where they're
  2013. 1:05:50valid because they're going to be
  2014. 1:05:51absorbed. Something happened. Unless of
  2015. 1:05:53course then it cascades into something
  2016. 1:05:55else, but that's a different kind of
  2017. 1:05:56trade. Now you're swing trading the
  2018. 1:05:58momentum to the downside, which is a
  2019. 1:06:00valid trade. You know, somebody can look
  2020. 1:06:03at the trade you take and say, you know,
  2021. 1:06:05that guy it's like, you know, the person
  2022. 1:06:08who's first is the person who gets shot
  2023. 1:06:10a lot of times, right? You don't want to
  2024. 1:06:11be the first guy through the door,
  2025. 1:06:12right? Uh so they'll they'll let you
  2026. 1:06:15take that trade. I'm going to let you
  2027. 1:06:17take that trade. I'll I'll tell you
  2028. 1:06:18that's the kind of trader I am. And then
  2029. 1:06:20when it breaks to the downside again,
  2030. 1:06:23I'm interested in that move. So you've
  2031. 1:06:25taken what I would consider to be a
  2032. 1:06:26higher risk trade. It's actually
  2033. 1:06:28probably a higher probability trade. You
  2034. 1:06:29probably make money on more of those,
  2035. 1:06:31but the way you make money on more of
  2036. 1:06:33those is by taking smaller profits,
  2037. 1:06:35which means you're not playing for a big
  2038. 1:06:36move. You're not playing for a long
  2039. 1:06:37move. So you
  2040. 1:06:40I don't think you get to say I just want
  2041. 1:06:42to hold these trades longer. How how can
  2042. 1:06:43I do that? Yeah, so I mean and I you
  2043. 1:06:46figured that out.
  2044. 1:06:46>> You you kind of spoken in labels of the
  2045. 1:06:49the principles of the market could be
  2046. 1:06:50mean reversion or trending or it could
  2047. 1:06:52be um momentum these orientations.
  2048. 1:06:56How do I know which one I am? Because
  2049. 1:06:58mean reverting and trend it depends on
  2050. 1:07:01the time frame you're looking at. I
  2051. 1:07:03could be trending daily bullish, but hey
  2052. 1:07:05H1 bearish, M15 bullish,
  2053. 1:07:07M1 bearish. Which time frame is the mean
  2054. 1:07:10reverting and trend?
  2055. 1:07:11>> Well, so you this is one very common
  2056. 1:07:15mistake. Uh
  2057. 1:07:18you have a time frame that you trade.
  2058. 1:07:19That's your trading time frame. And
  2059. 1:07:22let's say you get in let's just go back
  2060. 1:07:24to our example. Let's say you get short
  2061. 1:07:28on your one-minute time frame.
  2062. 1:07:30And then it turns into a loss and you're
  2063. 1:07:32holding it and I'm I'm guessing
  2064. 1:07:33viscerally you're like you wouldn't do
  2065. 1:07:35that. So like, you know, you you've
  2066. 1:07:37learned this. But let's say you do that
  2067. 1:07:38because you notice the 15-minute is
  2068. 1:07:41still kind of bearish.
  2069. 1:07:42And then that goes away on the 15-minute
  2070. 1:07:44and you you look at the oh, the hourly.
  2071. 1:07:47I have known more than one trader who
  2072. 1:07:51have entered trades on very low time
  2073. 1:07:54frames and ended up holding losing
  2074. 1:07:56trades for weeks.
  2075. 1:07:57The equivalent of risking a few hundred
  2076. 1:07:59dollars and losing tens of thousands of
  2077. 1:08:02dollars. This this is I call it the time
  2078. 1:08:04frame justification system.
  2079. 1:08:06And the way you avoid it is you say this
  2080. 1:08:09is the time frame I trade. I am in this
  2081. 1:08:12market because of this pattern, because
  2082. 1:08:14of this condition. And when that no
  2083. 1:08:16longer exists, I'm in business being in
  2084. 1:08:18that market. So you
  2085. 1:08:21you could look to develop a separate
  2086. 1:08:22edge on the daily time frame, but you
  2087. 1:08:26really need to stick with your time
  2088. 1:08:28frame. It also has to do
  2089. 1:08:30You want me to like
  2090. 1:08:31>> Go go go go.
  2091. 1:08:32>> It also has to do with how you manage
  2092. 1:08:33the trade because uh we could be in a
  2093. 1:08:36situation where, you know, let's say the
  2094. 1:08:39the the euro goes up and then comes down
  2095. 1:08:42and you you short it and I buy it at the
  2096. 1:08:45exact same time. We both can make money
  2097. 1:08:46on that trade.
  2098. 1:08:47>> Mhm.
  2099. 1:08:48>> You know, you you might be out and I
  2100. 1:08:51might buy more from you and then it goes
  2101. 1:08:53up later and you would have had a losing
  2102. 1:08:54trade, but you had a profitable trade
  2103. 1:08:56because you have your you took your
  2104. 1:08:57shorter profits as a mean-reverting
  2105. 1:08:59trader. Again, very broadly speaking,
  2106. 1:09:01needs to do. And because I was trading
  2107. 1:09:03momentum, I needed to be stickier and I
  2108. 1:09:05gave it more room to run. So it really
  2109. 1:09:07is it's not just what the market does,
  2110. 1:09:10it's you defining what you're going to
  2111. 1:09:12do in response to what the market does.
  2112. 1:09:13>> So you mentioned you should have a time
  2113. 1:09:14frame you select to trade upon.
  2114. 1:09:16Uh
  2115. 1:09:18by trade upon you mean a time frame I
  2116. 1:09:19execute upon? Because I can also do a
  2117. 1:09:22top-down analysis on multiple time
  2118. 1:09:24frames to ideate something and then I
  2119. 1:09:26will always execute on M1.
  2120. 1:09:28Uh should I separate those two things?
  2121. 1:09:30>> I mean you you already do. Where where
  2122. 1:09:32you get in trouble is if you're using
  2123. 1:09:34you know one common frame is you can use
  2124. 1:09:36several higher time frames or you can
  2125. 1:09:39use a higher and a lower time frame. But
  2126. 1:09:42and there are certainly things you can
  2127. 1:09:43do like you have your trading time frame
  2128. 1:09:46and you execute on the lower time frame
  2129. 1:09:48and then you manage it on your trading
  2130. 1:09:50time like though though but you you're
  2131. 1:09:52consistent. You know you're not it's not
  2132. 1:09:55like you're you're trading 15 5 and 1
  2133. 1:09:57minute and then all of a sudden I call
  2134. 1:09:59you up and you're trading on the 2-hour
  2135. 1:10:01chart today. That's a break of
  2136. 1:10:03discipline and that that will get you in
  2137. 1:10:05trouble. So yeah you just and you might
  2138. 1:10:07even you know I could imagine that you
  2139. 1:10:09tell me so most of my trades are on the
  2140. 1:10:111-minute chart but I have this one setup
  2141. 1:10:13that happens a few times a month on the
  2142. 1:10:14hourly chart and then I trade that. But
  2143. 1:10:16the trade is consistent on the time
  2144. 1:10:18frame that you're where you're getting
  2145. 1:10:20trouble is bouncing.
  2146. 1:10:20>> That makes sense to me. I I have with
  2147. 1:10:22time I've found with the style of
  2148. 1:10:24trading I have a top-down analysis
  2149. 1:10:26doesn't serve me because what relevance
  2150. 1:10:28does a monthly trend or a weekly or even
  2151. 1:10:31a daily trend have unless I'm in the
  2152. 1:10:34like the extremes
  2153. 1:10:35>> Yeah.
  2154. 1:10:35>> and most of the time I'm in no man's
  2155. 1:10:36land. How's that going to affect my
  2156. 1:10:37London session tomorrow? Uh so I've
  2157. 1:10:39avoided it and and I'm basically between
  2158. 1:10:41M15 and M1. Is there any merit in
  2159. 1:10:43thinking in this way?
  2160. 1:10:44>> That's where I've evolved over the
  2161. 1:10:45years. I most of my trading
  2162. 1:10:48So I got to be careful here because I
  2163. 1:10:49tell you most of my trading is on one
  2164. 1:10:51time frame but what I'm doing when I
  2165. 1:10:53look at that one time frame chart I'm
  2166. 1:10:55visualizing the higher time frame. So I
  2167. 1:10:57don't think there's anything wrong with
  2168. 1:10:58you seeing the structure on the higher
  2169. 1:10:59time frame. And I suppose you know
  2170. 1:11:02you're aware of levels on the higher
  2171. 1:11:03time frame that I may miss and if you're
  2172. 1:11:05coming into an important monthly level
  2173. 1:11:07that I might not see on my 1-minute
  2174. 1:11:09chart might be good to know that. Like
  2175. 1:11:10you know I I understand that argument.
  2176. 1:11:12But I think in general uh people add
  2177. 1:11:16pieces of complexity and it doesn't
  2178. 1:11:18necessarily add to to better results. I
  2179. 1:11:21I almost always have if I'm trading on
  2180. 1:11:23the daily time frame I'll have the
  2181. 1:11:24weekly time frame there but
  2182. 1:11:26I'm
  2183. 1:11:27my attention's on the daily. I I I'm
  2184. 1:11:29trading one chart, one time frame.
  2185. 1:11:31>> Back to that scenario we said where we
  2186. 1:11:33we probe a high and we're looking for
  2187. 1:11:34sellers to get in to then to then see a
  2188. 1:11:36rejection.
  2189. 1:11:37The word we used was liquidity. Do we
  2190. 1:11:39see sellers liquidity and and lack of
  2191. 1:11:41buyers liquidity?
  2192. 1:11:42I I've had this conversation with a few
  2193. 1:11:44people and they say to me, well,
  2194. 1:11:46there's liquidity everywhere in the
  2195. 1:11:47market literally every point, every pip,
  2196. 1:11:49every tick there is liquidity.
  2197. 1:11:50Will there be zones of heightened
  2198. 1:11:52liquidity such as a previous day's high?
  2199. 1:11:54And does heightened liquidity equal
  2200. 1:11:56better opportunity or does it kind of
  2201. 1:11:57also cancel out?
  2202. 1:12:01>> Giving
  2203. 1:12:02caveat of this answer, this is not my
  2204. 1:12:03specialization. I don't know for sure.
  2205. 1:12:06Yeah, there's liquidity at every point,
  2206. 1:12:07but there's a varying amount of
  2207. 1:12:09liquidity. What is liquidity? It's not
  2208. 1:12:11just resting orders. It's people being
  2209. 1:12:13willing to execute there and you
  2210. 1:12:14certainly could revisit a level very
  2211. 1:12:16quickly based on what price does and
  2212. 1:12:18liquidity is completely It's constantly
  2213. 1:12:19changing, right?
  2214. 1:12:21I think that ultimately
  2215. 1:12:23all of our edge comes from these shifts
  2216. 1:12:26in liquidity. That That's what it is.
  2217. 1:12:28And you know, then maybe we're back to
  2218. 1:12:30your order book question. Maybe people
  2219. 1:12:32are able to access that and understand
  2220. 1:12:34that in a way that I'm not really
  2221. 1:12:36looking at.
  2222. 1:12:37But I I think it all does come I mean,
  2223. 1:12:39we're using liquidity which sounds like
  2224. 1:12:41a big fancy word, but really what it
  2225. 1:12:43means is just people being willing to
  2226. 1:12:44transact at a price and people You know,
  2227. 1:12:46there's always a buy order for every
  2228. 1:12:48sell order, but you know, what urgency
  2229. 1:12:51is what what motivation is behind those?
  2230. 1:12:52That's really the question.
  2231. 1:12:54>> As you mentioned to yourself of you're
  2232. 1:12:55evolve ever evolving,
  2233. 1:12:57I feel like I've reached a a bit of a
  2234. 1:12:59ceiling and I know what the next step
  2235. 1:13:00would be. I just have a skill gap which
  2236. 1:13:03is
  2237. 1:13:04I do try to trade mechanical around
  2238. 1:13:06parameters and it is objective. I don't
  2239. 1:13:07trade intuition or gut feelings or
  2240. 1:13:10anything like that.
  2241. 1:13:11So therefore it would serve me to take
  2242. 1:13:12it algorithmic and and quantitative and
  2243. 1:13:15make it systematize and let's say even
  2244. 1:13:17eradicate human error by just having it
  2245. 1:13:19automated. I just don't really know how
  2246. 1:13:21to
  2247. 1:13:22you self-taught. So maybe you could shed
  2248. 1:13:24some light again especially in the AI
  2249. 1:13:25era how I myself and many others
  2250. 1:13:27listening could take that step.
  2251. 1:13:30>> Well, let me tell you a bad
  2252. 1:13:31>> thing to do with AI first. So this and
  2253. 1:13:33I'm seeing a lot of this. The bad thing
  2254. 1:13:35you do with AI is you go to your AI of
  2255. 1:13:39choice and you say what is the edge in
  2256. 1:13:42this pattern? Buy or find a
  2257. 1:13:45find a trading system that works for you
  2258. 1:13:48know what concept should I I don't think
  2259. 1:13:50at this point maybe this will change I
  2260. 1:13:51don't think AI is really good at that
  2261. 1:13:53ideation level.
  2262. 1:13:55I also would not trust an AI to say you
  2263. 1:13:59know we're talking the stats behind your
  2264. 1:14:01trade you know if the Euro takes out the
  2265. 1:14:03one minute at in this time window during
  2266. 1:14:05the day against these sessions
  2267. 1:14:08you know what are the stats after that?
  2268. 1:14:10I would if you asked AI I would have
  2269. 1:14:12zero confidence in that answer. What I
  2270. 1:14:14have done so I have a quantitative
  2271. 1:14:18framework that I've driven for years
  2272. 1:14:19that was you know written in MATLAB and
  2273. 1:14:22Python and R and I am now driving it
  2274. 1:14:26with AI. So I'm having conversations
  2275. 1:14:28with AI I'm asking these kinds of
  2276. 1:14:30questions but the AI is not answering
  2277. 1:14:33the questions the AI is driving my code
  2278. 1:14:36yeah but what it means is because I'm a
  2279. 1:14:38I'm a pretty shitty programmer to be
  2280. 1:14:40honest I I don't have to actually
  2281. 1:14:43program the system
  2282. 1:14:45figure out how to you know program the
  2283. 1:14:48levels.
  2284. 1:14:49I can tell the AI what I want it can
  2285. 1:14:51kick it back visual confirmation I can
  2286. 1:14:53look at a few hundred charts and say
  2287. 1:14:55yeah that 's right you know let's tweak
  2288. 1:14:57this and then I can run the test but
  2289. 1:14:59it's coming from my code driven by AI.
  2290. 1:15:02I think anybody can do something like
  2291. 1:15:04that but I think you have to kind of and
  2292. 1:15:06it sound you know I know it sounds
  2293. 1:15:07daunting but you can develop your own
  2294. 1:15:09framework that's going to be much more
  2295. 1:15:11valuable than asking AI for the answer.
  2296. 1:15:13>> I've spoken to a variety of guests on
  2297. 1:15:15the show and a unanimous common
  2298. 1:15:17denominator between all of them is the
  2299. 1:15:18emphasis they put on data and actually
  2300. 1:15:20knowing the inner workings and the
  2301. 1:15:22insights of your edge and your
  2302. 1:15:23performance. That 's why I'm proud to
  2303. 1:15:24bring a partner of the show TradeZella
  2304. 1:15:26the number one journaling back testing
  2305. 1:15:28and all-in-one insights experience
  2306. 1:15:30created by traders for traders. What
  2307. 1:15:32TradeZella really gives you is deep
  2308. 1:15:33insights about your trading that would
  2309. 1:15:35ordinarily not be visible, whether it's
  2310. 1:15:36through understanding your trade types
  2311. 1:15:38and playbooks or even insights powered
  2312. 1:15:40by artificial intelligence through Zella
  2313. 1:15:41AI. Whether you trade forex, futures,
  2314. 1:15:43cryptos, the stock market, it all
  2315. 1:15:45seamlessly connects to TradeZella, so
  2316. 1:15:47there is no additional work. You've seen
  2317. 1:15:48me reference it dozens of times and all
  2318. 1:15:50of the benefits I've had in my trading
  2319. 1:15:52from the insights I've found from my
  2320. 1:15:53TradeZella. So join myself and thousands
  2321. 1:15:55of other viewers of the show, you'll get
  2322. 1:15:57the best discount using the link in the
  2323. 1:15:58description or code TOT for Titans of
  2324. 1:16:01Tomorrow. So the reasons I would want to
  2325. 1:16:02go on this path and and I could hire
  2326. 1:16:05someone or whatever the answer would be.
  2327. 1:16:06But anyway, the reason I would do it is
  2328. 1:16:08to seek a certain advantage and the
  2329. 1:16:09advantage would be I guess an element of
  2330. 1:16:11buying back a bit of time.
  2331. 1:16:13The advantage would be when human error
  2332. 1:16:15comes in that could be eradicated. Even
  2333. 1:16:16to this day it does happen from time to
  2334. 1:16:18time.
  2335. 1:16:19And then also
  2336. 1:16:21performance related of like if I'm
  2337. 1:16:23traveling right now I just don't trade
  2338. 1:16:24because I'm in different time zones and
  2339. 1:16:25I can't really do it. But I could in
  2340. 1:16:27theory have it still running and check
  2341. 1:16:28up on upon it. So these would be the
  2342. 1:16:30benefits I would pursue.
  2343. 1:16:32But now you are also
  2344. 1:16:34algorithmic and mechanical and
  2345. 1:16:35objective, but also discretionary. And
  2346. 1:16:38it feels like then you would remove the
  2347. 1:16:40said benefits that I'm looking for by
  2348. 1:16:41being discretionary in in the mix.
  2349. 1:16:43>> So there's several ways to do it. One of
  2350. 1:16:45the classic ways to do it
  2351. 1:16:49would be to have and I think this could
  2352. 1:16:50be done pretty easily for what you're
  2353. 1:16:53looking to do.
  2354. 1:16:54Have your entry be automatic and then
  2355. 1:16:56you manage the exit.
  2356. 1:16:59So so the code takes the entry and then
  2357. 1:17:02you have a position. Now the problem is
  2358. 1:17:04if you're not at the screen, if you're
  2359. 1:17:05not paying attention, then
  2360. 1:17:08you know then then you have an issue.
  2361. 1:17:11The
  2362. 1:17:12there there are many ways to
  2363. 1:17:15intertwine discretion and systematic,
  2364. 1:17:18discretionary trading and systematic
  2365. 1:17:19trading. And I know most of the you talk
  2366. 1:17:21to systematic traders who tell you they
  2367. 1:17:23have no discretion.
  2368. 1:17:24That's not true because you still you
  2369. 1:17:27decided what to test. You decided this
  2370. 1:17:30is the word related to discretion,
  2371. 1:17:32right? You decided what to trade. You
  2372. 1:17:34decided what kind of risk to trade and
  2373. 1:17:36you're going to decide when the thing
  2374. 1:17:38stops working, when to turn it off. So
  2375. 1:17:40there is absolutely an element of
  2376. 1:17:42discretion. Uh, we can't have we we
  2377. 1:17:45can't be purely systematic. You can't be
  2378. 1:17:47pure Maybe someone can. I've never seen
  2379. 1:17:49it. A purely discretionary trader would
  2380. 1:17:52have none of what we're discussing. They
  2381. 1:17:53would have no plan. They would look at
  2382. 1:17:55the chart. They say I'm going to
  2383. 1:17:56sell it there, buy it there, and then
  2384. 1:17:58you know they're just making
  2385. 1:18:00That person over time, they're going to
  2386. 1:18:02accumulate experience, and they're going
  2387. 1:18:04to accumulate a system around their If
  2388. 1:18:08they're successful, they're going to
  2389. 1:18:09have some systematic elements to their
  2390. 1:18:11to their discretion.
  2391. 1:18:13I
  2392. 1:18:14would suspect for you the if
  2393. 1:18:18it does not remove the part of you being
  2394. 1:18:20at the screen, but I also don't think
  2395. 1:18:22you can trade the time frame you're
  2396. 1:18:23going to trade without you being at the
  2397. 1:18:25screen. Uh, but that idea about having
  2398. 1:18:28the entry and then you managing the
  2399. 1:18:30exits and of course when we program the
  2400. 1:18:32entry, we'll also program some exits,
  2401. 1:18:34right? So one of the things that we need
  2402. 1:18:38to do whenever we start meddling with a
  2403. 1:18:40systematic approach, we need to analyze
  2404. 1:18:43our discretionary inputs and see is it
  2405. 1:18:45adding anything. And I've seen this. You
  2406. 1:18:47know, I mean I've I've traded some
  2407. 1:18:49systematic approaches in the past, and I
  2408. 1:18:51found on balance my discretionary input
  2409. 1:18:54is actually helpful, but I'm monitoring
  2410. 1:18:56it, and I know, and I can I can quantify
  2411. 1:18:59the alpha contribution of my discretion
  2412. 1:19:03over just letting the system run. Also
  2413. 1:19:05has to do with the kinds of systems. If
  2414. 1:19:08we're trading a system that's based
  2415. 1:19:09around the kinds of patterns that we're
  2416. 1:19:11talking about, it would be somewhat
  2417. 1:19:13similar to discretionary trader.
  2418. 1:19:15I think I have valid input as a
  2419. 1:19:17discretionary trader, but the the weird
  2420. 1:19:20averaged alpha approach, I probably
  2421. 1:19:22don't have any valid discretionary input
  2422. 1:19:24there, I would assume.
  2423. 1:19:25>> Is time of day of execution relevant to
  2424. 1:19:27you?
  2425. 1:19:28>> Yeah, I mean it this goes back to
  2426. 1:19:30liquidity, right? Because uh
  2427. 1:19:32the most liquid market can be completely
  2428. 1:19:36dull and dead for hours and hours out of
  2429. 1:19:38the day. Uh there's certainly are
  2430. 1:19:40intraday edges that are very time frame
  2431. 1:19:43specific.
  2432. 1:19:44Uh you can
  2433. 1:19:46if people want to start doing
  2434. 1:19:47quantitative research, you can you you
  2435. 1:19:50can start chopping the day up into hours
  2436. 1:19:52or half hours and asking, you know, hey,
  2437. 1:19:54what if we make a new high in this half
  2438. 1:19:56hour? Is that different than making it?
  2439. 1:19:57And the answer is yes. So it it matters
  2440. 1:20:00quite a lot. And this is why I'm sure
  2441. 1:20:02you've talked to day traders who only
  2442. 1:20:04trade a few hours a day. They've carved
  2443. 1:20:06their edge and you know I mean maybe
  2444. 1:20:08this is you too. You've carved your edge
  2445. 1:20:09out of a specific window when specific
  2446. 1:20:11things are happening and absolutely that
  2447. 1:20:13matters.
  2448. 1:20:14>> Yeah, in the end trading 24 hours a day
  2449. 1:20:16would be chaotic. So it ends up being 2
  2450. 1:20:18hours a day for me. But I don't know if
  2451. 1:20:20I found anything specific that would
  2452. 1:20:22benefit me or because there are times
  2453. 1:20:23that I I'm seeing the exact same pattern
  2454. 1:20:25if if I'm to call it. Signatures of
  2455. 1:20:27price that I'm interested in happening
  2456. 1:20:28at a London Open, I take it. Happening 3
  2457. 1:20:31hours later in my no man's
  2458. 1:20:33no man's land time time frame,
  2459. 1:20:35I won't take it. And I don't know if I
  2460. 1:20:37improve the odds because I just I'm
  2461. 1:20:39susceptible to less opportunities,
  2462. 1:20:41therefore less losses, or is that
  2463. 1:20:43actually a heightened statistical skew
  2464. 1:20:44in my 1 2 hour window that I have?
  2465. 1:20:46>> I would think that could be tested. I
  2466. 1:20:48mean if you want to if you want to send
  2467. 1:20:49me some I can I can throw it into my
  2468. 1:20:51research process and take a look at it.
  2469. 1:20:52>> on that offer. That sounds fun. Um in
  2470. 1:20:54fact, what would be the best way for me
  2471. 1:20:56to approach all of this if I want to
  2472. 1:20:58become a little bit more quantitative?
  2473. 1:21:00Would it be
  2474. 1:21:01upload price action to to something and
  2475. 1:21:03then just start conversing through AI?
  2476. 1:21:05Or would it be here's my last 100
  2477. 1:21:06trades, feed it in and then ask
  2478. 1:21:08questions?
  2479. 1:21:09>> I think I think both of those are valid
  2480. 1:21:11approaches. I think also there's
  2481. 1:21:12something to be said for
  2482. 1:21:15constructing what I would consider to be
  2483. 1:21:17kind of
  2484. 1:21:19a
  2485. 1:21:22I don't have a word for it.
  2486. 1:21:24So we're talking
  2487. 1:21:26the kind of quantitative analysis we're
  2488. 1:21:27talking about is one thing. There's also
  2489. 1:21:28a way which we haven't really discussed
  2490. 1:21:30of coming with a specific rule set,
  2491. 1:21:32right? So um you know I 'm looking for
  2492. 1:21:36price to be below the 200 day moving
  2493. 1:21:37average and then to cross above the the
  2494. 1:21:39five the three crosses above the day
  2495. 1:21:42average and I take a buy with a stock.
  2496. 1:21:44You know, so you and you've talked a lot
  2497. 1:21:46of people probably who do this. That's
  2498. 1:21:47certainly valid way to
  2499. 1:21:50to solve the puzzle.
  2500. 1:21:52I think there's some value in doing that
  2501. 1:21:54because you will learn how so I would
  2502. 1:21:57try to I would try to structure a system
  2503. 1:21:59like that.
  2504. 1:22:01And you know
  2505. 1:22:02I I gave one example, but for you it
  2506. 1:22:05might be the rule might be in this time
  2507. 1:22:07window
  2508. 1:22:09if the market goes above this level and
  2509. 1:22:11is back below it and I don't if you back
  2510. 1:22:13below it in 1 minute or you know
  2511. 1:22:14whatever that's that's a variable you
  2512. 1:22:16can look at. You know, I'll get short
  2513. 1:22:18here. My stop will be at or beyond that
  2514. 1:22:21previous high. My profit target is here.
  2515. 1:22:24If I don't have a profit in 15 minutes I
  2516. 1:22:26get out of the you know and those are
  2517. 1:22:28all variables and things. There there
  2518. 1:22:29would be a lot of value for you, I
  2519. 1:22:32think, in taking what you're doing on a
  2520. 1:22:34discretionary basis and putting it into
  2521. 1:22:37a clean rule set like that and then
  2522. 1:22:39seeing it run and then going back and
  2523. 1:22:42you know so one of the thing it's very
  2524. 1:22:43easy to push the button and let the
  2525. 1:22:45system run. What you also need to do, I
  2526. 1:22:48think this this is how this is how I
  2527. 1:22:49learned a lot is go back and look at the
  2528. 1:22:51specific executions. And you got to look
  2529. 1:22:54at a lot of them. You you know you need
  2530. 1:22:56to you
  2531. 1:22:58don't go and look at three trades and
  2532. 1:22:59say well those looked stupid. I wouldn't
  2533. 1:23:01have done those. You know we're talking
  2534. 1:23:03dozens of trades and you know if you've
  2535. 1:23:05done a back test hundreds it takes a lot
  2536. 1:23:07of time. You have to be
  2537. 1:23:09a not normal person to to to want to do
  2538. 1:23:11that. And if you are normal it will make
  2539. 1:23:13you a not normal.
  2540. 1:23:15But you know you're going through and
  2541. 1:23:16you're looking at hundreds of these and
  2542. 1:23:17you and you can take some notes, but
  2543. 1:23:19you're also building some this is how
  2544. 1:23:21how we build intuition is seeing
  2545. 1:23:23patterns, right? And seeing occurrences.
  2546. 1:23:26And then you'll think about making them
  2547. 1:23:28my I've had cases where I've tried to
  2548. 1:23:30systemize approaches exactly what I'm
  2549. 1:23:33explaining here and I was not
  2550. 1:23:34successful. I was doing something on a
  2551. 1:23:37discretionary basis and I went through
  2552. 1:23:39iterations of this
  2553. 1:23:41or
  2554. 1:23:42approach with it
  2555. 1:23:43couldn't could not systemize, but it
  2556. 1:23:45made my discretionary trading better
  2557. 1:23:47because it made me look at a lot of
  2558. 1:23:48patterns. It made me look at a lot of
  2559. 1:23:50outcomes. It made me think about what
  2560. 1:23:52exactly was contributing to the edge.
  2561. 1:23:54And, you know, so the system development
  2562. 1:23:56was a failure, but I made money from it.
  2563. 1:23:58And that's that's like having a making
  2564. 1:23:59money on a losing trade.
  2565. 1:24:01>> Exactly.
  2566. 1:24:01>> good.
  2567. 1:24:02>> I'm curious if overfitting, which is I
  2568. 1:24:04mean, yesterday's conversation I had, it
  2569. 1:24:05seemed like a big topic.
  2570. 1:24:07>> It's a big topic.
  2571. 1:24:08>> But I I don't know if it's like a
  2572. 1:24:09problem that most people won't encounter
  2573. 1:24:11or is something actually problematic.
  2574. 1:24:13>> It's a problem in that kind of in that
  2575. 1:24:16kind of development very specific
  2576. 1:24:18because uh
  2577. 1:24:20what are what are the variables? Okay,
  2578. 1:24:21so we're talking about a time window.
  2579. 1:24:24We're talking about a certain
  2580. 1:24:25penetration Sorry. Certain penetration
  2581. 1:24:27through the high.
  2582. 1:24:28Uh we're talking about where you put
  2583. 1:24:30your profit target and then and there's
  2584. 1:24:31others. But
  2585. 1:24:33what you could do is test every
  2586. 1:24:34combination of those.
  2587. 1:24:35And
  2588. 1:24:37some of those will work.
  2589. 1:24:39You know, if we're testing random levels
  2590. 1:24:42back to our Fibonacci discussion, some
  2591. 1:24:44of those random levels are going to
  2592. 1:24:45work. There there there's going to be
  2593. 1:24:47non-random action. You may know, you
  2594. 1:24:49know, statistically if you do a bunch of
  2595. 1:24:51tests, you can't just look at the
  2596. 1:24:52p-value. You have to adjust for the fact
  2597. 1:24:54I did a bunch of tests. I did 20 tests.
  2598. 1:24:56Therefore, there's a probability at a
  2599. 1:24:5895% that one of those is going to look
  2600. 1:24:59good, right? So you you have to I think
  2601. 1:25:03this is where a lot of those kinds of
  2602. 1:25:04system developers get in trouble is
  2603. 1:25:07they'll do all of these tests and then
  2604. 1:25:09it is massively overfitted. And there
  2605. 1:25:11are ways to overcome that, right? Like
  2606. 1:25:13if I see if you tell me you tested this
  2607. 1:25:15grid of five variables and you found
  2608. 1:25:18some magical combination that worked and
  2609. 1:25:20all the others look like garbage and one
  2610. 1:25:23works, I don't think you have anything.
  2611. 1:25:25But if there's kind of a plateau where a
  2612. 1:25:27lot of things work, you might have
  2613. 1:25:29something. The gold standard is to do
  2614. 1:25:33your development and hold out an out of
  2615. 1:25:35sample and then test that. Most people
  2616. 1:25:37don't do that right because the problem
  2617. 1:25:39with out of sample is how many times can
  2618. 1:25:40you do the out of sample?
  2619. 1:25:44So if you're doing your system
  2620. 1:25:46development and then you're like, let me
  2621. 1:25:47just look at this on my you you've
  2622. 1:25:49completely invalidated your test. Most
  2623. 1:25:50people the outside of you know real
  2624. 1:25:54quant firms don't have the discipline to
  2625. 1:25:56do all of the system development which
  2626. 1:25:58may take weeks and weeks and weeks and
  2627. 1:26:00then to run it on the out of sample
  2628. 1:26:01which answers to some extent what if I
  2629. 1:26:04had traded this over the past year, six
  2630. 1:26:06months or whatever if you see
  2631. 1:26:08persistence there. You can do a cheap
  2632. 1:26:10out of sample
  2633. 1:26:12uh all of my tests I also look at like
  2634. 1:26:15uh I call them temporal quantize. So you
  2635. 1:26:17know I'm I 'm slicing whatever data that
  2636. 1:26:20I'm looking at into periods and I want
  2637. 1:26:23to see some persistence. If it only
  2638. 1:26:24looks good if I'm testing 10 years and
  2639. 1:26:27only looks good in two and a half of
  2640. 1:26:28those it's probably not great.
  2641. 1:26:31>> We also spoke about the beginning of
  2642. 1:26:32alpha decay. So if I'm going through all
  2643. 1:26:34this in sample out of sample testing
  2644. 1:26:36and then I start to deploy it is there
  2645. 1:26:38any degradation rate you found of of
  2646. 1:26:41alpha decaying and therefore an edge an
  2647. 1:26:43edge has a certain shelf life?
  2648. 1:26:45>> I think so. Yeah.
  2649. 1:26:46>> What what is that
  2650. 1:26:47you know
  2651. 1:26:47>> It it varies but it's it's varies it's
  2652. 1:26:49real and it seems to go in one direction
  2653. 1:26:52and it's not it's not the direction you
  2654. 1:26:53want. It's it's not the answer you want.
  2655. 1:26:55Uh I think you can greatly accelerate
  2656. 1:26:58alpha decay by publishing.
  2657. 1:27:00>> Mm.
  2658. 1:27:00>> So tell people about it. Uh
  2659. 1:27:02one of the
  2660. 1:27:04absolute standards is you know there's
  2661. 1:27:07tons of academic papers that have real
  2662. 1:27:08alphas published.
  2663. 1:27:10Uh there are books that have real and
  2664. 1:27:12there's also lots of books that don't
  2665. 1:27:13have alphas. They say they do but they
  2666. 1:27:15don't. Uh but what you will find is that
  2667. 1:27:18those alphas that were good at one point
  2668. 1:27:20in time
  2669. 1:27:21invariably they they've degraded since
  2670. 1:27:23publication.
  2671. 1:27:25And this goes back to our discussion
  2672. 1:27:26about this doesn't mean that you can't
  2673. 1:27:29that I can't trade the pullback that you
  2674. 1:27:31can't trade the tests through that level
  2675. 1:27:34but how we trade it today may degrade
  2676. 1:27:37and we may need to modify that. You know
  2677. 1:27:39you may need to change your change what
  2678. 1:27:42the penetration looks like, change what
  2679. 1:27:43the action after you know the the any of
  2680. 1:27:45these variables may need some adjustment
  2681. 1:27:48and I think as a discretionary trader,
  2682. 1:27:52when you're in touch with the data,
  2683. 1:27:54you're going to do some of those
  2684. 1:27:55adjustments automatically. And that goes
  2685. 1:27:57back to where I said, you know, I tell
  2686. 1:27:59you I've traded the same way for
  2687. 1:28:00decades. It's a lie because I modified
  2688. 1:28:03the subtle things around it, sometimes
  2689. 1:28:05without even knowing. You know, I I
  2690. 1:28:07don't even know I'm doing something
  2691. 1:28:08different until I go back and look at a
  2692. 1:28:10different period. I look, oh, my stops
  2693. 1:28:11are in quite a different place now. And
  2694. 1:28:14you know, I evolved with the market. And
  2695. 1:28:15that's the advantage of discretionary
  2696. 1:28:17trading.
  2697. 1:28:17>> Right at the beginning, on your own
  2698. 1:28:18admission, you said your intuition sucks
  2699. 1:28:20and in fact I should do the opposite of
  2700. 1:28:21your intuition and I would make money.
  2701. 1:28:23Um
  2702. 1:28:24you you mentioned intuition is kind of
  2703. 1:28:26like your subconscious picking up on
  2704. 1:28:27patterns.
  2705. 1:28:28>> Maybe.
  2706. 1:28:28>> Uh so that feels like some sort of
  2707. 1:28:31advantage. But anyhow, the question is
  2708. 1:28:33if you are optimizing through quant
  2709. 1:28:36through probabilities and you have a
  2710. 1:28:37quantitative approach and you say your
  2711. 1:28:39intuition sucks, what's the purpose of
  2712. 1:28:41discretion?
  2713. 1:28:43>> I
  2714. 1:28:45there are patterns that I can trade. And
  2715. 1:28:48I don't I don't think it's the same
  2716. 1:28:49thing as intuition. I think if I
  2717. 1:28:51identify a pattern like the pattern
  2718. 1:28:53you've described to me,
  2719. 1:28:55if I can identify that pattern on a
  2720. 1:28:56chart, I might not be able to quantify
  2721. 1:28:58it.
  2722. 1:28:59I might not be able to explain it to
  2723. 1:29:01code in a way that I could execute, but
  2724. 1:29:04I can see it on a chart, I can trade it
  2725. 1:29:07on a chart. In fact, it might be quite
  2726. 1:29:09easy for me, but if I describe it in
  2727. 1:29:11code, what I'm going to find is I'll
  2728. 1:29:12describe that, but I'll also describe
  2729. 1:29:15four examples that aren't actually the
  2730. 1:29:17thing. So there is some degree of
  2731. 1:29:19filtering. Maybe if I were a better
  2732. 1:29:21programmer, maybe you know, some of the
  2733. 1:29:24modeling tools now I did quite a bit of
  2734. 1:29:27work many, many years ago with
  2735. 1:29:30was working with a group of of
  2736. 1:29:32engineers who were looking at the
  2737. 1:29:34algorithms that radars use to a radar
  2738. 1:29:37from an airplane looking down at another
  2739. 1:29:40airplane,
  2740. 1:29:41everything's just reflected back to the
  2741. 1:29:43radar. You you may know this or you
  2742. 1:29:45many of your listeners do, uh you get a
  2743. 1:29:47lot of ground clutter. So it's not the
  2744. 1:29:50simplest thing to pull. So there are
  2745. 1:29:52algorithms, of course, that process
  2746. 1:29:54that. So they were looking at those
  2747. 1:29:56kinds of algorithms, which also had
  2748. 1:29:57found use in speech recognition, and
  2749. 1:30:00they were trying to apply those kinds of
  2750. 1:30:02algorithms to my discretionary pattern
  2751. 1:30:04recognition.
  2752. 1:30:05It's an interesting project. We spent a
  2753. 1:30:06lot of time on it.
  2754. 1:30:08Zero Zero result from it.
  2755. 1:30:11I think, you know, maybe some of the
  2756. 1:30:12modeling tools are a little bit more
  2757. 1:30:14sophisticated now. Maybe you could show
  2758. 1:30:17the right kind of training environment.
  2759. 1:30:19This is a pattern. This is not a
  2760. 1:30:21pattern. And then it will see things. It
  2761. 1:30:23could maybe tell us things about our
  2762. 1:30:24perception. We don't know.
  2763. 1:30:25>> And I feel like we could have another
  2764. 1:30:27like 10-hour conversation. So I think
  2765. 1:30:29the best way for me to end is asking
  2766. 1:30:31you,
  2767. 1:30:32is there anything I should have asked
  2768. 1:30:33you today?
  2769. 1:30:36>> I think we did a really good job to pat
  2770. 1:30:38ourselves on the back staying focused
  2771. 1:30:40because there were a lot of things that
  2772. 1:30:41we could have talked about and there are
  2773. 1:30:42I think that
  2774. 1:30:44one of the
  2775. 1:30:46things that is occupying a lot of my
  2776. 1:30:48thought at this point in my life is
  2777. 1:30:51thinking about life outside of markets
  2778. 1:30:53and thinking about, you know, how do you
  2779. 1:30:57how do you interact with the markets
  2780. 1:30:59responsibly with the degree let's say
  2781. 1:31:02focus because it's a nice word, but we
  2782. 1:31:04really mean single-minded crazy
  2783. 1:31:06obsession
  2784. 1:31:07while also being a good human and also
  2785. 1:31:10enjoying the time that we're alive. And
  2786. 1:31:12I think that crosses into many other
  2787. 1:31:15questions that we can talk about
  2788. 1:31:16meditation and spiritual practice and
  2789. 1:31:18art and the meaning and the meaning of
  2790. 1:31:20everything and
  2791. 1:31:22I mean we could talk about knowledge and
  2792. 1:31:24you know what we know and how we know it
  2793. 1:31:26and how sure we are with I think that's
  2794. 1:31:28actually
  2795. 1:31:29set maybe that's a separate interview
  2796. 1:31:31sometime. I think we did a good job
  2797. 1:31:32staying on track today.
  2798. 1:31:33>> Yeah, it was a very colorful
  2799. 1:31:34conversation with a lot of wisdom and
  2800. 1:31:35lessons. I thoroughly enjoyed it. Thank
  2801. 1:31:37you for the opportunity.
  2802. 1:31:37>> I enjoyed it too. Thank you.

About this transcript

This page contains the full transcript of The Quant Who Studied 700 Yrs Of Price Data: Most Levels Are Noise! THESE Are The 2 Levels To Watch! by Titans Of Tomorrow, generated from the public captions YouTube serves with the video. The transcript has 18,623 words across 2,802 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.