YouTube2Text

Evolving Trends: How Quants Are Earning Alpha · Rishi Narang — Transcript

by Chat With Traders · 9,954 words · 1,730 segments · language en · Watch on YouTube

Full transcript

  1. 0:02Markets, speculation, and risk. This is
  2. 0:06the Chat with Traders podcast, hosted by
  3. 0:08Aaron Fifield.
  4. 0:11Hey traders, welcome. Thank you for
  5. 0:13joining me on episode 207.
  6. 0:17This week I had the pleasure of catching
  7. 0:19up with Rishi Narang. Longtime listeners
  8. 0:22of the show may recall I interviewed
  9. 0:24Rishi back in 2016
  10. 0:27on episode 54.
  11. 0:30By the way, Rishi's the founding
  12. 0:32principal of T2AM,
  13. 0:35a fund of funds investment manager with
  14. 0:37a sole focus of allocating to
  15. 0:39quantitative trading strategies. The
  16. 0:42fund was formed in 2005.
  17. 0:45The Rishi's time in markets extends more
  18. 0:47than 20 years as he co-founded
  19. 0:50distinguished quant shop Tradeworks in
  20. 0:531999.
  21. 0:55Rishi's also the author of an excellent
  22. 0:58book, Inside the Black Box, and an
  23. 1:01advisor to DARPA's financial markets
  24. 1:04vulnerabilities project.
  25. 1:07Yes, that is DARPA as in the government
  26. 1:09agency responsible for the robot dogs
  27. 1:12you've probably seen in those
  28. 1:14somewhat terrifying videos.
  29. 1:17So, the purpose of our catch-up was
  30. 1:20mostly to go over prominent trends he's
  31. 1:22seen develop in recent times specific to
  32. 1:25quant trading and investing.
  33. 1:28Without giving away too much, it
  34. 1:30involves talk of stat arb and index
  35. 1:33rebal strategies,
  36. 1:35an attraction to China's market,
  37. 1:38narrower trading models, use cases of
  38. 1:41alt data, and we wrap up discussing the
  39. 1:44rise of retail or DIY quants.
  40. 1:48I can appreciate these topics won't be
  41. 1:50of interest to everyone,
  42. 1:52but as I've said before, I think it is
  43. 1:54highly beneficial to have some insight
  44. 1:56on how the larger players are moving in
  45. 1:59markets and how they are earning alpha.
  46. 2:02Finally, please excuse the odd bird
  47. 2:05sound. I believe Rishi was wandering
  48. 2:07outside his home as we were talking.
  49. 2:10And after this first question, you
  50. 2:12actually won't hear from me for about 20
  51. 2:14minutes or so. My suggestion, think of
  52. 2:17this first part as more of a talk or a
  53. 2:20an interesting presentation from Rishi,
  54. 2:23and then an interview follows. Please
  55. 2:26enjoy. I present to you from Los
  56. 2:28Angeles, California, Rishi Narang.
  57. 2:34You know, speaking here with regard to
  58. 2:36research and strategy,
  59. 2:39what has stood out to you as notable
  60. 2:40trends in the field of quantitative
  61. 2:42trading and investing?
  62. 2:45Sure. Yeah, so there's there's a couple
  63. 2:47of few things to highlight. One One is
  64. 2:50the
  65. 2:51Well, I mean, there's a few things
  66. 2:52actually. So,
  67. 2:53one is that as the landscape has
  68. 2:55shifted, meaning
  69. 2:58you try to earn alpha in a context. It's
  70. 3:01not something that happens in a vacuum.
  71. 3:03And that context is sort of maybe best
  72. 3:06described as follows. Alpha comes from
  73. 3:10inefficiencies in asset prices
  74. 3:13that are themselves caused by either
  75. 3:17inefficiencies or relative indifferences
  76. 3:21on the part of various market
  77. 3:22participants.
  78. 3:24So, an old example would be that
  79. 3:27institutional traders, institutional
  80. 3:29investors were pretty sloppy in their
  81. 3:31handling of
  82. 3:33trading of single stocks.
  83. 3:35And so, they would do these large lumpy
  84. 3:37orders and and ignore market impact.
  85. 3:41And that basically created a whole game
  86. 3:44called stat arb, statistical arbitrage.
  87. 3:47And what would happen is there would be
  88. 3:49a price dislocation in some name.
  89. 3:53You would see that, but it's hard to
  90. 3:56know
  91. 3:57from the outside whether that came for
  92. 3:59good reason or not.
  93. 4:00And so, rather than just buy a name that
  94. 4:03that fell or short a name that went up a
  95. 4:05bunch,
  96. 4:07you would find something that reduces
  97. 4:09your risk.
  98. 4:10And so, you'd put on a hedge. So, you
  99. 4:13could imagine hedging with the index.
  100. 4:16But the problem with hedging with the
  101. 4:18index is there's a lot of basis risk.
  102. 4:20You know, if you're
  103. 4:21if you're buying Tesla and shorting the
  104. 4:23S&P 500, there's a lot of idiosyncratic
  105. 4:25risk in Tesla.
  106. 4:28Well, I maybe picked a really crappy
  107. 4:29example because I don't know who a good
  108. 4:31peer for Tesla is that's publicly listed
  109. 4:33today, but
  110. 4:34the point is that you would try to find
  111. 4:36something that reduces your risk. So, at
  112. 4:38least an automaker reduces your risk,
  113. 4:40right? So, maybe Ford or General Motors
  114. 4:43is a better hedge for Tesla
  115. 4:45than
  116. 4:46you know, the S&P 500 index. So,
  117. 4:50that game of
  118. 4:53finding related things and shorting the
  119. 4:56outperformer and buying the
  120. 4:57underperformer, which is also known as
  121. 4:59statistical arbitrage or relative mean
  122. 5:00reversion,
  123. 5:02existed in the '80s, '90s, early 2000s
  124. 5:05because of
  125. 5:07the sloppiness of a certain class of
  126. 5:08investors,
  127. 5:10mostly for that reason.
  128. 5:12Well, as those folks evolved,
  129. 5:15that inefficiency kind of went away.
  130. 5:17And it changed. So, relative mean
  131. 5:19reversion still exists, but it exists
  132. 5:21for a very different reason, and that
  133. 5:23reason also causes it to manifest in a
  134. 5:25different way. Which means if you're
  135. 5:26pursuing statistical arbitrage today,
  136. 5:29your model better not look just like it
  137. 5:31did
  138. 5:32in, say, 1997.
  139. 5:34Because otherwise, you're like not going
  140. 5:36to make any money and probably just lose
  141. 5:38money.
  142. 5:39There are a number of ways in which the
  143. 5:40landscape has shifted, really more than
  144. 5:41we have time to get into now, but you
  145. 5:43can just think about it for yourself how
  146. 5:45much the world has changed, say, in the
  147. 5:46last 15-20 years.
  148. 5:48There's been a gigantic move from
  149. 5:51passive to active,
  150. 5:52other way around. Wow. From active to
  151. 5:54passive, sorry.
  152. 5:56And a gigantic move from
  153. 6:00discretionary trading
  154. 6:03interaction with the markets to
  155. 6:06algorithmic interaction with the
  156. 6:07markets.
  157. 6:09At least within US equities, for
  158. 6:10example, almost all the volume is well,
  159. 6:13not almost all, a giant portion of the
  160. 6:16volume
  161. 6:17is now done market on close as opposed
  162. 6:20to throughout the day.
  163. 6:22And so, there's a bunch of these
  164. 6:23dynamics that
  165. 6:25are all intersecting with one another
  166. 6:27and changing the approach
  167. 6:30to making alpha.
  168. 6:32So, that's one thing that
  169. 6:35that we've seen that's changing the game
  170. 6:38around around us and
  171. 6:40forcing us to adapt and evolve.
  172. 6:43So, one example is
  173. 6:46the
  174. 6:47relative increase in prominence of index
  175. 6:50rebalance strategies. So, these are
  176. 6:52strategies that
  177. 6:53try to anticipate what names are going
  178. 6:55to be added to various indices.
  179. 6:58Why are we doing that? Well,
  180. 7:00or added and deleted, by the way. Why
  181. 7:03are we doing that? Well, we're doing
  182. 7:05that because there's
  183. 7:06a ton of money
  184. 7:08that gets invested on the back of what,
  185. 7:10say, Standard & Poor's puts in the S&P
  186. 7:12500.
  187. 7:14And that's true for MSCI indices and all
  188. 7:16kinds of other indices as well.
  189. 7:18So, all this passive money isn't exactly
  190. 7:21passive. It's just sort of sheep-like
  191. 7:24following
  192. 7:26the the the
  193. 7:29dictums.
  194. 7:30I guess that would be dicta
  195. 7:32of various index construction
  196. 7:34companies.
  197. 7:36And so, those index construction
  198. 7:38companies
  199. 7:39say, "Okay,
  200. 7:41we're including company X in the S&P
  201. 7:44500."
  202. 7:45And that announcement
  203. 7:47means that now all of the Vanguards and
  204. 7:49Fidelities and all of the other index
  205. 7:52managers and the the folks who
  206. 7:54administer the SPY ETF
  207. 7:57and so on, they all have to go out and
  208. 7:58buy company X in the right proportion.
  209. 8:02And it of course, with a fixed number of
  210. 8:05names type of index like the S&P 500, if
  211. 8:08a name is in, then a name is also out.
  212. 8:10And so, those folks then also have to
  213. 8:11sell
  214. 8:12that other name.
  215. 8:14So, that's interesting, too, right? So,
  216. 8:16you have something where there's going
  217. 8:17to be enormous buying pressure and
  218. 8:18enormous selling pressure.
  219. 8:20Pretty useful to know about that in
  220. 8:22advance if you could.
  221. 8:25You kind of can because a lot of them
  222. 8:27publish
  223. 8:28indicators of how or guide guidebooks or
  224. 8:32guidelines for how they think about
  225. 8:34adding and removing names for inclusion.
  226. 8:37So,
  227. 8:39that game
  228. 8:40really exists because of inefficiencies
  229. 8:42in the way passive folks invest. There's
  230. 8:44an indifference there.
  231. 8:46You know, if you're the S&P 500, you
  232. 8:49only care a little bit about the
  233. 8:50performance of the S&P 500. You're
  234. 8:52you know, certainly for an ETF manager,
  235. 8:55you don't really care.
  236. 8:56Cuz
  237. 8:58by definition, you get those stocks in
  238. 9:00those proportions and you just sit and
  239. 9:02hold them.
  240. 9:04Guess what return you're going to get?
  241. 9:05You're going to get the S&P 500 return.
  242. 9:07If you have adverse market impact,
  243. 9:10meh, it doesn't really matter. I mean,
  244. 9:12it goes into the price and whatever.
  245. 9:14So, there's not that much incentive to
  246. 9:16be that clever about
  247. 9:19managing market impact.
  248. 9:21I mean, there is to some extent, but
  249. 9:23it's limited.
  250. 9:24And so,
  251. 9:25there are a whole bunch of strategies
  252. 9:27that exist.
  253. 9:28Not a whole bunch, but there's a whole
  254. 9:30bunch of money being managed around this
  255. 9:32strategy of index rebalancing.
  256. 9:34And it's not just the S&P, of course.
  257. 9:36It's all these indices globally. So,
  258. 9:37that's one big example of how the world
  259. 9:39has changed is this active to passive
  260. 9:42thing. And by the way, even if you're a
  261. 9:43stat arb manager,
  262. 9:45you now kind of need to start accounting
  263. 9:47for index rebals in your price action.
  264. 9:51So, as an example,
  265. 9:53if your
  266. 9:54model is ignorant of the fact
  267. 9:57that company X has just been added and
  268. 9:59company Y has just been deleted from the
  269. 10:00index.
  270. 10:02And you're just watching price action
  271. 10:04relative to peers,
  272. 10:06then suddenly company X is soaring and
  273. 10:09company Y is reeling.
  274. 10:11And you may get in front of that
  275. 10:13mountain of flow that's coming your way.
  276. 10:16So, you kind of have to start to have
  277. 10:18almost a risk factor
  278. 10:20or a
  279. 10:22uh an adverse selection avoidance
  280. 10:24technique
  281. 10:25for index rebal. So, even if you're not
  282. 10:28trading it as an alpha, which you would
  283. 10:30do if you're doing index rebal, you
  284. 10:32still have to have index rebal
  285. 10:34if you're going to try to be
  286. 10:36good at the game.
  287. 10:38So, that's one thing. Another thing is
  288. 10:40uh like a big
  289. 10:42dispersion or um diaspora of
  290. 10:46the places and markets that quant
  291. 10:48strategies are are applied in. So, you
  292. 10:51know, Indian stat arb and emerging
  293. 10:53market stat arb and China stat arb China
  294. 10:56stat arb is like a huge business right
  295. 10:57now.
  296. 10:58Uh I don't even know the numbers
  297. 11:00offhand, but the amount of money going
  298. 11:02into both onshore money and offshore
  299. 11:05money going into quantitative
  300. 11:08equities and futures trading
  301. 11:10of mainland Chinese assets is just
  302. 11:12stupendous.
  303. 11:14Billions and billions and billions of
  304. 11:15dollars. Huge number.
  305. 11:17Um and the number of firms is is also
  306. 11:20pretty big. You know, there are a lot of
  307. 11:22folks who, for example,
  308. 11:24worked at US firms and
  309. 11:27then realized, well, there's this more
  310. 11:29or less frontier market where there's a
  311. 11:31lot more alpha, a lot more retail
  312. 11:32participation,
  313. 11:34and a lot more inefficiencies in China.
  314. 11:37It's also harder in a bunch of ways to
  315. 11:39trade offshore
  316. 11:41or if you're an offshore entity,
  317. 11:42especially up until quite recently.
  318. 11:45And so,
  319. 11:47maybe, you know, if you were someone who
  320. 11:48was here on H-1B visa in the US
  321. 11:52and worked for Two Sigma or D. Shaw or
  322. 11:56whatever quant firm,
  323. 11:58maybe just go home and set up a shop
  324. 12:01there
  325. 12:02uh in China and have a bunch of local
  326. 12:04advantages over the US and and UK firms
  327. 12:08that are trying to play that game.
  328. 12:10So, that's definitely been happening. Um
  329. 12:12is Oh, and and the application of quant
  330. 12:14techniques into all kinds of other
  331. 12:15areas, by the way.
  332. 12:17Venture capital,
  333. 12:18uh crypto,
  334. 12:20credit, peer-to-peer lending,
  335. 12:23um municipal bond market making. I mean,
  336. 12:26all kinds of seen so many interesting
  337. 12:28things that are not just
  338. 12:30basic equity stat arb in the developed
  339. 12:32markets or basic futures strategies in
  340. 12:35the developed markets.
  341. 12:37And I guess the third big trend
  342. 12:40is maybe the most interesting,
  343. 12:43which is the
  344. 12:46there's been a shift in the balance
  345. 12:48between taking over-fitting risk and
  346. 12:50under-fitting risk.
  347. 12:52And just philosophically, I think it's
  348. 12:54really interesting.
  349. 12:56In the old days, well, pretty dirty
  350. 13:00insult for a quant was that they've
  351. 13:01overfit.
  352. 13:03But now we're seeing folks do a variety
  353. 13:06of things
  354. 13:07that
  355. 13:09their actions are telling you that
  356. 13:11they're at least as afraid of
  357. 13:12under-fitting as of over-fitting.
  358. 13:15That looks like a couple different
  359. 13:16things. So, first of all, in the old
  360. 13:19days, for the purposes of largely
  361. 13:20cross-validation and statistical
  362. 13:22significance,
  363. 13:23models were typically universal, meaning
  364. 13:26for some universe of many securities,
  365. 13:30basically without exclusion,
  366. 13:32uh
  367. 13:33this model applies to all of them. So,
  368. 13:35if there are
  369. 13:37uh all US stocks, but then
  370. 13:39realistically, you can only trade the
  371. 13:40top, say, 3,000 of them if if that's
  372. 13:43what you determine from a liquidity
  373. 13:44standpoint,
  374. 13:46then you have one model with one set of
  375. 13:47parameters,
  376. 13:49one set of factors,
  377. 13:51uh and one set of weights for all 3,000
  378. 13:54stocks, meaning the drivers of your
  379. 13:56forecast for Walmart uh are the same as
  380. 14:00for Pfizer, are the same as for Amazon,
  381. 14:03are the same as for Google and Facebook,
  382. 14:06and Tesla. All those companies are
  383. 14:09forecasted with the exact same set of
  384. 14:11factors, the same weights, the same
  385. 14:12parameters.
  386. 14:15Um we're now seeing industry-specific
  387. 14:18and even company-specific models, or if
  388. 14:20you're talking about futures, individual
  389. 14:22market models.
  390. 14:24And what you're seeing with those is
  391. 14:28a decision that your prior
  392. 14:32might be a lot stronger on a single
  393. 14:34company forecast forecasting model.
  394. 14:37Like, I can know what the KPRs KPIs are
  395. 14:40for some specific company.
  396. 14:43But um
  397. 14:45I might not have a lot of data history
  398. 14:48to prove that out in a backtest, but I'm
  399. 14:50willing to trust my prior more
  400. 14:53uh and not worry as much about having
  401. 14:55statistical significance.
  402. 14:58Uh another easy example would be
  403. 15:00something like inventory data. So, you
  404. 15:02know, on US companies' quarterly
  405. 15:05reports, there's a standard line item
  406. 15:07for inventory. Well,
  407. 15:09if you're a bank, what is inventory?
  408. 15:12If you're a
  409. 15:14uh you know, an
  410. 15:16an audit and consulting firm like Price
  411. 15:18Waterhouse Coopers or Ernst & Young or
  412. 15:20whatever,
  413. 15:21what what is that? You know?
  414. 15:25Uh what is inventory? But if you're a
  415. 15:28a manufacturer of television sets, well,
  416. 15:30now we know what inventory is, right?
  417. 15:33So, but that is important information
  418. 15:35for the latter kind of company and not
  419. 15:36for the former.
  420. 15:38And so,
  421. 15:39you don't want to throw out that data
  422. 15:41potentially.
  423. 15:42Uh but that does mean you have a
  424. 15:44different model for those industries
  425. 15:46where
  426. 15:48inventory is a thing and a different
  427. 15:50model for those sectors and industries
  428. 15:52where it's not.
  429. 15:53So, that's one expression of this shift
  430. 15:57in the balance between over- and
  431. 15:58under-fitting risk.
  432. 16:00The second is
  433. 16:02the increased use of alternative data
  434. 16:05sources.
  435. 16:06And this is related to the first thing,
  436. 16:10which we'll call narrower models as
  437. 16:12opposed to more broad models.
  438. 16:14Um models are applied in a applied in a
  439. 16:16more narrow set of um of securities. For
  440. 16:20the second thing,
  441. 16:21um this is now models that are driven by
  442. 16:24data that are only available for certain
  443. 16:27kinds of companies. So, a great example
  444. 16:29is uh is credit card data. So, I'm
  445. 16:33guessing that your listeners will know
  446. 16:35that their credit card spending is being
  447. 16:39tracked by various companies,
  448. 16:42and whether individualized or
  449. 16:45aggregated, that data is being sold
  450. 16:47to hedge funds,
  451. 16:50uh who use it mostly quantitatively, but
  452. 16:53not always, uh to make decisions about
  453. 16:56what to buy and sell. And basically,
  454. 16:58what they're doing is
  455. 17:00looking for a faster or more accurate
  456. 17:04estimation of, say, revenues for
  457. 17:07uh certain kinds of companies. Well, if
  458. 17:09it's credit card data,
  459. 17:11it's interesting cuz it's you can't buy
  460. 17:13everything and you don't buy everything
  461. 17:14on credit cards. And in any case,
  462. 17:16there's only a subset of things
  463. 17:19that go into
  464. 17:22uh the category of like
  465. 17:25things that humans, like regular people,
  466. 17:27buy. So, as an example, it doesn't tell
  467. 17:29you anything about uh how Boeing is
  468. 17:32doing on its aerospace sales. Like,
  469. 17:34credit card data does not apply to
  470. 17:36Boeing.
  471. 17:37Probably doesn't apply much to auto
  472. 17:39manufacturers, either, cuz most people
  473. 17:41are not buying cars on their credit
  474. 17:43card.
  475. 17:44They're either sending a wire or writing
  476. 17:46a check or writing a check and then
  477. 17:48having a financing line.
  478. 17:50And that's a different thing.
  479. 17:52It tells you a lot about what they're
  480. 17:54doing with Amazon
  481. 17:56and what they're doing with Macy's and
  482. 17:59whatever else. Like, so it's really
  483. 18:01useful for retail and consumer and
  484. 18:03hospitality-type companies.
  485. 18:05Um and a few other categories, but, you
  486. 18:08know, credit card data is really
  487. 18:09interesting for
  488. 18:11a few hundred companies in the US.
  489. 18:14Not very interesting for the other
  490. 18:17two and a half, three and a half
  491. 18:18thousand easily tradable companies.
  492. 18:22So, that's the second thing is this sort
  493. 18:24of trend towards alternative data,
  494. 18:26which is very much hand in hand with the
  495. 18:30trend towards narrower models.
  496. 18:33The third thing is
  497. 18:36the increased use of machine learning
  498. 18:39or artificial intelligence uh
  499. 18:41strategies. So, that's the other really
  500. 18:43big piece in all of this.
  501. 18:46So, uh you know, these things, by
  502. 18:48definition,
  503. 18:49are looking at data
  504. 18:52to tell you how what the forecasting
  505. 18:56variables should be and how much they
  506. 18:57should matter.
  507. 18:59Uh and we can talk about the different
  508. 19:01use cases of ML and AI if you want, but
  509. 19:03um there are some firms
  510. 19:06that claim to be making forecasts
  511. 19:07without priors, without saying, "Oh, I
  512. 19:09have this economic model of how
  513. 19:10companies'
  514. 19:11asset prices work, what drives a stock
  515. 19:14to go up, what drives a stock to go
  516. 19:15down.
  517. 19:17I'm just going to let the machine figure
  518. 19:18it out."
  519. 19:19And that's inherently a a more fitting
  520. 19:23kind of an exercise.
  521. 19:25So, here you've
  522. 19:26obviously outlined three different
  523. 19:31trends that you've that you've noticed
  524. 19:33uh in space of quant trading and
  525. 19:35investing. Going right back to the
  526. 19:37start, and I just have a few questions
  527. 19:39about each of these.
  528. 19:41You spoke about knowing in advance
  529. 19:44when a symbol may be added to an index.
  530. 19:49Can you maybe go into that a little bit
  531. 19:50more about how you may know in advance.
  532. 19:54Sure. Yeah, so
  533. 19:56in many cases and this will just be one
  534. 19:58example for the sake of time.
  535. 20:00In many cases most of the determination
  536. 20:03is simply a market capitalization based
  537. 20:05cutoff.
  538. 20:06So, we're going to take like the Russell
  539. 20:093000 Russell 2000.
  540. 20:11There's a certain ranking of market caps
  541. 20:14that happens.
  542. 20:16And if it's
  543. 20:17company 1001
  544. 20:20through 3000
  545. 20:22in terms of the ranking by market cap
  546. 20:26kind of presumably
  547. 20:28that company is supposed to be in the
  548. 20:31Russell 2000.
  549. 20:33And they have certain dates that they
  550. 20:35announce these um inclusions and
  551. 20:37exclusions. They announce the new
  552. 20:39composition of the index and then dates
  553. 20:42at which those changes are made
  554. 20:44effective. And there's some time in
  555. 20:46between there. So, there's a range of
  556. 20:48strategies that you can do. Similar to
  557. 20:51the way that people treat merger arb,
  558. 20:52you can
  559. 20:54look at like the history of all mergers
  560. 20:56and say, "Well, what kinds of companies
  561. 20:59get bought?"
  562. 21:00And then I'm going to go out and screen
  563. 21:02the world for those kinds of companies.
  564. 21:04And maybe buy some or all of them.
  565. 21:07And
  566. 21:09then some of them are going to be
  567. 21:11targets for for takeovers.
  568. 21:13So, there you're trying to get an ahead
  569. 21:15of the
  570. 21:17uh the acquirer. Here you're trying to
  571. 21:19get ahead of the market cap
  572. 21:22um based index
  573. 21:24solution or index provider. And then all
  574. 21:27of the people and firms that have to
  575. 21:30follow suit and either bring in the new
  576. 21:32names or get rid of the old ones.
  577. 21:34It's not exactly front running, but all
  578. 21:36alpha is in a sense probabilistic front
  579. 21:39running. You're trying to know before it
  580. 21:41happens
  581. 21:42that other people are going to buy this
  582. 21:44stock in aggregate or sell this stock or
  583. 21:46this asset
  584. 21:48in in advance. Um
  585. 21:50And you're doing that with a bunch of
  586. 21:51hard work and not with any kind of
  587. 21:53actual knowledge of their trades.
  588. 21:55But you do have to get there before they
  589. 21:57do. Otherwise, it's not alpha. Then it's
  590. 22:01you know, something else that doesn't
  591. 22:03work, you know.
  592. 22:04I wonder what the name for that is.
  593. 22:06Doing uh doing all the right trades, but
  594. 22:09doing them after the fact. I mean, I
  595. 22:11guess that's trend following, but
  596. 22:14This information, I mean, is that
  597. 22:16publicly available to everybody? Yeah.
  598. 22:19The information about what's going to be
  599. 22:21included into an index? Some of it is.
  600. 22:23So, as an example, um the index
  601. 22:26providers have gotten smart about as as
  602. 22:29all capitalist companies generally are,
  603. 22:31but how to extract money for everything.
  604. 22:34So, just as exchanges are now charging a
  605. 22:36huge amount for data
  606. 22:38that
  607. 22:39you know, is a major part of their
  608. 22:41revenue streams
  609. 22:42even though they're really not in the
  610. 22:44business of selling data, they're in the
  611. 22:45business of
  612. 22:47processing transactions of securities
  613. 22:50and providing a safe environment for the
  614. 22:51transaction of securities. And so, they
  615. 22:53charge fees for that, too. But then they
  616. 22:56they know you have to have that data.
  617. 22:58If you're like a high frequency trader,
  618. 23:00you have to have the data of every
  619. 23:01single exchange
  620. 23:03for various reasons in the US. And so,
  621. 23:07um
  622. 23:08those exchanges have you know, extracted
  623. 23:10a toll for for that data. Well, just so
  624. 23:12the index providers are
  625. 23:14charging
  626. 23:15for index composition data.
  627. 23:17And they are
  628. 23:20and some of it's not cheap. I mean, to
  629. 23:22get like Russell, S&P, MSCI
  630. 23:26that is some hundreds of thousands of
  631. 23:28dollars a year.
  632. 23:30Just for those three and not even for
  633. 23:32all regions.
  634. 23:35Wow, really? Mhm. That much?
  635. 23:37anyone like it's not like you have to
  636. 23:39know somebody who knows somebody to get
  637. 23:40the data, but you do have to have
  638. 23:42hundreds of thousands of dollars of data
  639. 23:43of budget for the data. Yeah.
  640. 23:46That's true for a lot of stuff, man.
  641. 23:48And this you're talking about the data
  642. 23:50around index rebalancing. Specifically
  643. 23:53that, yes. And all the No, no, just to
  644. 23:55be clear, this is not index rebal data.
  645. 23:57This is index composition data. This is
  646. 23:59to tell you just what is in the S&P 500
  647. 24:02right now. What tickers in at what
  648. 24:04weights. That's it. That's all that
  649. 24:06hundreds of thousands of dollars gets
  650. 24:07you.
  651. 24:08But you can imagine if you're doing
  652. 24:10index rebal,
  653. 24:11it's kind of an important ingredient in
  654. 24:14in doing the trade right.
  655. 24:16So, what what are some of the trades you
  656. 24:18see around this? Like is it I mean, I
  657. 24:22presume that they're probably a little
  658. 24:23bit more sophisticated than just buying
  659. 24:25a symbol which you think is going to be
  660. 24:28uh added into an index.
  661. 24:31Whoa. Why would you presume that? It's
  662. 24:33That's it. That's the trade.
  663. 24:36Now, the sophistication may be in why do
  664. 24:39you think that it's going to be added?
  665. 24:41But um and there are a range of
  666. 24:43strategies around that. So, there are
  667. 24:45folks who traffic in
  668. 24:47like the lower probability names in
  669. 24:49index rebal versus those that traffic in
  670. 24:52the names that have already been
  671. 24:53announced, but it hasn't completely been
  672. 24:56um implemented yet.
  673. 24:59So, there's like a range of certainty
  674. 25:01about whether or not this name is going
  675. 25:03to be included. And as a result of that
  676. 25:06range of certainty, there are different
  677. 25:08players with different risk appetites
  678. 25:10and different models. So, let's say for
  679. 25:12example,
  680. 25:13that you're talking about the Russell
  681. 25:151000, which if I if memory serves and if
  682. 25:18my knowledge is correct, is simply the
  683. 25:201000 largest market caps in the US
  684. 25:23equity market.
  685. 25:25So, now let's say that um there's a
  686. 25:27rebal coming up, an announcement coming
  687. 25:29up about the rebal.
  688. 25:31Uh and when and when I say coming up,
  689. 25:33like how far in advance you trade is an
  690. 25:35open question here. So, but let's just
  691. 25:37pretend that the answer is 3 months
  692. 25:39before the rebal, I'm going to start
  693. 25:41putting on positions.
  694. 25:43Okay?
  695. 25:44So, it's uh I think their rebals are
  696. 25:45announced in March, but I don't
  697. 25:47remember. Could be November. I have no
  698. 25:49idea. Whatever. I don't remember the
  699. 25:50frequency. But let's say it's March. So,
  700. 25:52I'm going to start putting on my
  701. 25:53positions in December, January.
  702. 25:56At that time,
  703. 25:57there's a bunch of companies
  704. 26:00that are say ranked from 900 or 800 to
  705. 26:061200 in market cap, which over the
  706. 26:10course of 3 months
  707. 26:13might move up or down enough in market
  708. 26:15cap to either get them excluded from the
  709. 26:17top 1000 or included in the top 1000.
  710. 26:19So, this is interesting cuz this isn't
  711. 26:21just about is this company going to be
  712. 26:22good?
  713. 26:23It's is it going to be good at It's is
  714. 26:25it going to be better or worse
  715. 26:27than the companies around it in this
  716. 26:29specific weird peer group of companies
  717. 26:31that may have nothing to do with each
  718. 26:32other otherwise.
  719. 26:34But they're all not really competing,
  720. 26:35but you can think of them as competing
  721. 26:37like horse racing
  722. 26:39to be in that top 1000 so that they get
  723. 26:41included. Now again, maybe the companies
  724. 26:43don't care if they're included in the
  725. 26:44Russell 1000 or not, but their stock
  726. 26:46prices care.
  727. 26:47Now let's say that I'm really good at
  728. 26:49forecasting 3 months price outcomes for
  729. 26:51companies. Well, I better use that
  730. 26:53information. And then I'll
  731. 26:57have a little bit more conviction to buy
  732. 26:59bigger
  733. 27:01those names that I'm going to hold for 3
  734. 27:03months-ish and are likely to be included
  735. 27:06cuz then when that announcement comes,
  736. 27:07there's a big pop.
  737. 27:09And as the certainty builds, there's a
  738. 27:11big pop. So, there's this kind of weird
  739. 27:14short-term momentum thing that happens
  740. 27:16in those in that category of names that
  741. 27:19has nothing to do with say the top 10
  742. 27:21companies in the Russell 1000 cuz
  743. 27:23there's almost a zero probability
  744. 27:25that in 3 months those companies are uh
  745. 27:29are going to drop out of the index.
  746. 27:31Right? So, it's a little bit like
  747. 27:33options theory. If you're the number one
  748. 27:35ranked market cap in the US or in any
  749. 27:38market,
  750. 27:39and you know, you're far and away the
  751. 27:41biggest, which happens, it would be like
  752. 27:45um what is the value of a deep
  753. 27:48out of the money put
  754. 27:49on that company
  755. 27:51with a 3 month expiry. The value of that
  756. 27:53is like very close to zero.
  757. 27:55It's basically zero.
  758. 27:59If you're talking about something where
  759. 28:01it's a 3 month expiry for very near the
  760. 28:04money options,
  761. 28:05well, now that's fun, right? And if you
  762. 28:08have information that gives you a clue,
  763. 28:10then you should use that information.
  764. 28:13Does that make sense?
  765. 28:15It does make sense, yeah.
  766. 28:17And what's more, there's a
  767. 28:20there's an asymmetry here.
  768. 28:23If you're company number 999,
  769. 28:26the last time or 1000, the last time the
  770. 28:29Russell 1000 was constituted,
  771. 28:32and you're still kind of 999, 1000,
  772. 28:36there's a chance you get excluded, which
  773. 28:38means there's a whole ton of money that
  774. 28:39has to come out of you.
  775. 28:42If you're company 1001,
  776. 28:46and you remain company 1001, there's not
  777. 28:48a lot of downside for that price because
  778. 28:51it's already excluded and it's still
  779. 28:52excluded.
  780. 28:54So, for those companies that are outside
  781. 28:56the index looking in,
  782. 28:58there is more upside than downside. And
  783. 29:00for those companies that are barely in
  784. 29:02the index, there's more downside than
  785. 29:03upside. So, it's just very interesting
  786. 29:05that dynamic and what it means when
  787. 29:08you're building a model.
  788. 29:11So, ultimately you're trying to
  789. 29:13buy early enough predicting that a
  790. 29:16company may be added to an index um in
  791. 29:19order to essentially
  792. 29:21front run all the flow which is going to
  793. 29:22be coming into that stock over the
  794. 29:24coming weeks and months.
  795. 29:26Yeah, and again, this word front run has
  796. 29:28like actual legal meaning and we don't
  797. 29:31mean it that way, but yes.
  798. 29:32Of course.
  799. 29:33it that. You want to get in front of it.
  800. 29:35Which is a lot like front running, but
  801. 29:37but the legal definition of front
  802. 29:38running is about, you know, having
  803. 29:39knowledge of someone like actual
  804. 29:41knowledge of someone's order and then
  805. 29:42going in front of it. Like if you call
  806. 29:44your broker and say, "I I to buy a
  807. 29:45billion shares of IBM.
  808. 29:47And then your broker's like, "Cool. Let
  809. 29:50me go buy a
  810. 29:51million shares of IBM right before that,
  811. 29:54and then I'll do your order." Because
  812. 29:56that way I'll profit from all your
  813. 29:57market impact. Let's just say you're
  814. 29:59you're trying to be the first to act.
  815. 30:01Yeah, or or yeah, certainly not the
  816. 30:03last. You're trying to be early to act.
  817. 30:05Early, yeah.
  818. 30:07Yeah.
  819. 30:07The second point you made was about
  820. 30:10China, which I thought was quite
  821. 30:12interesting. You said there's a lot of
  822. 30:13startup opportunities there right now.
  823. 30:17Um how come that is the case?
  824. 30:20It's a combination of factors. So,
  825. 30:22there's onshore money
  826. 30:24and offshore money. There's onshore
  827. 30:26firms and offshore firms. Like you can't
  828. 30:30have onshore money
  829. 30:31if you're an offshore firm.
  830. 30:33That's like not a thing.
  831. 30:35Uh you can have offshore money if you're
  832. 30:38an onshore firm. So, there's a supply
  833. 30:41thing and a demand thing that's
  834. 30:42happening. From the supply side, the
  835. 30:45ability to
  836. 30:47trade
  837. 30:49China as an offshore investor, that's
  838. 30:50recently been a little bit um that's
  839. 30:53been made a little bit easier recently.
  840. 30:55And so that's that's one thing, right?
  841. 30:57So, if you're like Two Sigma or PDT or
  842. 31:01some other US or whatever non-Chinese
  843. 31:03firm, you can now access China
  844. 31:06relatively easily actually. There's a
  845. 31:08bunch of like international brokers like
  846. 31:11Morgan Stanley and UBS and so on that
  847. 31:14can give you access to that market uh
  848. 31:16simply very simply. Um and there are
  849. 31:19rules around shorting and in shorting
  850. 31:21individual names and shorting the index
  851. 31:23and so on. So, you can even be market
  852. 31:26neutral.
  853. 31:28If I understand correctly, and I haven't
  854. 31:29studied this too too much, but um if I
  855. 31:32understand correctly, if you're an
  856. 31:33onshore
  857. 31:35firm managing onshore money,
  858. 31:37uh you cannot short
  859. 31:40individual names.
  860. 31:42I'm not even sure if you can short the
  861. 31:44index, but maybe you can.
  862. 31:46But that's not Chinese money.
  863. 31:48Um that's been going on a little bit of
  864. 31:50time,
  865. 31:51but there's also a different supply
  866. 31:53dynamic there,
  867. 31:55which is that you
  868. 31:57uh as I said, have a lot of new supply
  869. 32:00of quants who used to work at
  870. 32:04non-Chinese firms
  871. 32:06and have decided, "Let's set up shop in
  872. 32:08China, where we live, where there's a a
  873. 32:10less efficient market."
  874. 32:12That second thing, where there's a less
  875. 32:14efficient market, that statement
  876. 32:16is what's driving the demand side of
  877. 32:19this. Which is to say, again, as some of
  878. 32:22the more obvious and straightforward
  879. 32:25markets to access
  880. 32:26and bigger markets have filled up. If
  881. 32:28you look at China, it's a really big
  882. 32:30market. It's a really big economy, only
  883. 32:32set to get bigger.
  884. 32:34And, you know, there's a rush to get in
  885. 32:37there before it gets too crowded and be
  886. 32:40early and gather data and all those
  887. 32:42things that give you an early mover
  888. 32:44advantage.
  889. 32:45Um access to talent, access to data,
  890. 32:48storing your own data is super valuable,
  891. 32:51but this you know, you can't do that
  892. 32:52until you start doing it. So, like go
  893. 32:54start doing it. So, that there's been
  894. 32:56like a lot of that.
  895. 32:58Have you allocated to any managers who
  896. 33:01uh trading Chinese markets, like in
  897. 33:04order for for your fund to have exposure
  898. 33:06there?
  899. 33:07Yeah, we have offshore two offshore
  900. 33:10Chinese well,
  901. 33:12money with two firms that are both based
  902. 33:13in mainland China, but are offering
  903. 33:16offshore strategies to offshore
  904. 33:18investors.
  905. 33:19And yeah, we've we've picked a couple of
  906. 33:21them and we've looked at dozens.
  907. 33:23Okay.
  908. 33:24Is that just a recent thing?
  909. 33:27Pretty recent. Um
  910. 33:29both this year. Uh
  911. 33:31I want to say the first one
  912. 33:33went live mid-year, but we've probably
  913. 33:34you know, that means we've been talking
  914. 33:35to them for a little while. Yeah, okay.
  915. 33:39You also brought up the point about uh I
  916. 33:41might get this slightly wrong, but it
  917. 33:43was something to the extent of things
  918. 33:46which are used in quant being applied to
  919. 33:50other areas outside of financial
  920. 33:52markets. I know you mentioned venture
  921. 33:54capital, and there were a couple other
  922. 33:55things as well. Could you speak to that
  923. 33:58a little bit more?
  924. 33:59Yeah, and it's again the same drive for
  925. 34:01diversification and and access to
  926. 34:04markets that are a little weirder and
  927. 34:06more complicated to to So, there's some
  928. 34:09more barriers to entry. And if you think
  929. 34:11about this like just from a basic
  930. 34:12economic principle standpoint, it's the
  931. 34:13same thing as saying, "Well, there are
  932. 34:15barriers to entry to making this kind of
  933. 34:17widget,
  934. 34:18but if I happen to have some edge in
  935. 34:19making that widget, then I should
  936. 34:21probably go do that because I'm going to
  937. 34:22enjoy a competitive advantage for a
  938. 34:25while."
  939. 34:26So, and being early is its own
  940. 34:28competitive advantage because that means
  941. 34:29I have a head start on others.
  942. 34:32But you know, there's something you have
  943. 34:33to risk. You have to put up a bunch of
  944. 34:34upfront costs to get there. So, credit
  945. 34:37um
  946. 34:38you know, this is like debt issued by
  947. 34:40companies that
  948. 34:41uh that are So, it's not government
  949. 34:42debt, it's usually uh you know, it's
  950. 34:45it's it's debt from private companies.
  951. 34:48That debt trades.
  952. 34:50So, it's not just that the company
  953. 34:52issues
  954. 34:53debt and you decide, "I will lend to
  955. 34:55this company."
  956. 34:56You can buy and sell that debt. And
  957. 34:58there is a market for it.
  958. 35:00It's in the US
  959. 35:02corporate credit market, not
  960. 35:04um fully exchange uh traded, and in
  961. 35:07fact, it's mostly not.
  962. 35:09Um but, you know, some of the credit
  963. 35:11default swaps are exchange-traded, and
  964. 35:13it's a changing market, an evolving
  965. 35:14marketplace.
  966. 35:16Um
  967. 35:17it it has very different challenges, and
  968. 35:19one of the most obvious is is data.
  969. 35:22Um like it's not that complicated to
  970. 35:24think about what would cause a bond to
  971. 35:26be more attractive relatively and what
  972. 35:28would cause it to be less attractive.
  973. 35:30What's complicated is asking
  974. 35:33like,
  975. 35:34"How do I even track the history of
  976. 35:37bonds?" So, unlike stocks, bonds expire.
  977. 35:40And on top of that, they're issued at
  978. 35:41various coupons at various times and
  979. 35:44various durations. And so,
  980. 35:46managing that data set, it's like it's
  981. 35:48just a different problem
  982. 35:50from uh managing normal equity data set
  983. 35:53or even futures data set. You know,
  984. 35:55futures expire too, but there's far
  985. 35:56fewer of them
  986. 35:58than there are companies and credits.
  987. 36:01So, yeah, there's like this crazy data
  988. 36:03wrangling problem. There's systematic
  989. 36:06strategies
  990. 36:07uh in peer-to-peer lending. So, this is
  991. 36:09like Lending Club and Lending Tree,
  992. 36:11where if you're just some random person
  993. 36:14and you want to borrow $5,000 for
  994. 36:16whatever random reason,
  995. 36:18you can apply for a loan,
  996. 36:21and other random people who have 5,000
  997. 36:24extra dollars sitting around can lend
  998. 36:25that to you and earn on it like their
  999. 36:28local savings account, but then they're
  1000. 36:29taking credit risk on you as an
  1001. 36:32individual, the random borrower. And so,
  1002. 36:35um there are
  1003. 36:36now quant strategies out there
  1004. 36:39that are pouring over credit data and
  1005. 36:41pouring over characteristics of
  1006. 36:43borrowers
  1007. 36:44and saying, "Well,
  1008. 36:46this borrower might not have the
  1009. 36:47greatest credit score." I don't know if
  1010. 36:48you guys have the equivalent in
  1011. 36:50Australia or elsewhere, wherever
  1012. 36:52listeners may be, but in the US, you
  1013. 36:54know, you have a credit score that a
  1014. 36:56credit agency uh has a list of factors
  1015. 36:59that
  1016. 37:00uh describe your creditworthiness. And,
  1017. 37:04you know, like a in a very popular such
  1018. 37:06scheme, a really high credit score, I
  1019. 37:08think the maximum is 850.
  1020. 37:10850. I don't know why it's not a
  1021. 37:12thousand or why it's not a hundred, but
  1022. 37:14it's 850.
  1023. 37:15So, that's what it is.
  1024. 37:16And um you know, you're considered as
  1025. 37:19having good credit, I think, if your
  1026. 37:21credit's above 750, let's call it. And
  1027. 37:24so, you know, if you're a an 820 credit
  1028. 37:27score person, it's very easy to get
  1029. 37:29money fairly cheaply. If you're a 600
  1030. 37:31credit score person, it's a lot less
  1031. 37:33easy. But it might be that that's just
  1032. 37:35because of something stupid. Because
  1033. 37:37these credit these credit ratings are
  1034. 37:39not like
  1035. 37:41God telling you this is a good lender,
  1036. 37:43or good borrower, I mean. This is like a
  1037. 37:45a flawed universally applied model
  1038. 37:47telling you that. Maybe some random
  1039. 37:49circumstance happened three years ago
  1040. 37:51beat up your credit score, but it turns
  1041. 37:53out you're actually a really good
  1042. 37:54borrower who pays everything back.
  1043. 37:56So, they look for those kinds of people
  1044. 37:59to try to find the ones that
  1045. 38:01are better borrowers.
  1046. 38:04And but because their credit score
  1047. 38:06sucks, they know they're going to have a
  1048. 38:08higher interest rate.
  1049. 38:10Does that make sense? You try to find
  1050. 38:11like these good value loans to make,
  1051. 38:14where you get a high yield, but the risk
  1052. 38:16isn't that bad.
  1053. 38:18It's not not in line how high the yield
  1054. 38:20is.
  1055. 38:21Um so, yeah, there's systematic
  1056. 38:22strategies for that, and you can tell
  1057. 38:23there's a whole different set of data
  1058. 38:25there and a whole different kind of
  1059. 38:26problem there than there is when you're
  1060. 38:28trying to forecast the price of IBM.
  1061. 38:32Gotcha. That's really quite fascinating.
  1062. 38:37And some of this stuff isn't new. Some
  1063. 38:39of it's been going on for a while.
  1064. 38:41Um you know, systematic credit trading's
  1065. 38:43been going on
  1066. 38:4510, 15 years now.
  1067. 38:47But at the same time, the prominence of
  1068. 38:49it is much higher. Like there's at least
  1069. 38:51two
  1070. 38:52stand-alone quant funds
  1071. 38:54that are doing just that.
  1072. 38:57And then the magnitude of that activity
  1073. 38:59within some of the larger multi-strats
  1074. 39:01is much larger than it used to be. And
  1075. 39:03it's the same for quant trading in
  1076. 39:05China. Like that's been happening for a
  1077. 39:06minute, but it's just it's now like on
  1078. 39:09the map, you know.
  1079. 39:11So, the third point
  1080. 39:14uh I wanted to ask you a question on
  1081. 39:16that uh with regard to underfitting and
  1082. 39:19overfitting a model. Do you have any
  1083. 39:21tips for how to get the balance right?
  1084. 39:25Oof, that's actually quite legitimately
  1085. 39:28really hard. I mean, there's so many
  1086. 39:29factors.
  1087. 39:31And
  1088. 39:32the reason that this change is happening
  1089. 39:35is once again both from a supply
  1090. 39:38perspective and from a demand
  1091. 39:40perspective. And I know I'm mapping
  1092. 39:42things to supply and demand. This isn't
  1093. 39:44like a clean mapping, but I'll explain
  1094. 39:46what I mean. So, on the supply side,
  1095. 39:48we have an increasing number of people
  1096. 39:51who are
  1097. 39:52in jobs in quant firms
  1098. 39:55that have training in machine learning
  1099. 39:56and artificial intelligence. And some of
  1100. 39:58those folks used to work at places like
  1101. 40:01Facebook, and they're just sick of
  1102. 40:03selling advertisements to people.
  1103. 40:05Right? Like they thought, "Oh, cool. I
  1104. 40:07get to work for this really cool
  1105. 40:08company.
  1106. 40:09It's a tech company. I'm not selling my
  1107. 40:11soul. Look, it's social media. I'm like
  1108. 40:13connecting families and friends and
  1109. 40:15former colleagues and alumni."
  1110. 40:18But then it turns out what what I'm
  1111. 40:19really doing is I'm selling people ads.
  1112. 40:21And that gets at some point maybe
  1113. 40:22grating on your soul.
  1114. 40:24And so, you quit that job and you go
  1115. 40:25work for a different soul-crushing place
  1116. 40:27that's just trying to make rich people
  1117. 40:28richer. Cool.
  1118. 40:30Nonetheless,
  1119. 40:31um that is one
  1120. 40:34um
  1121. 40:35source of increased supply. The other is
  1122. 40:37the field itself has advanced. I mean,
  1123. 40:39the problem of forecasting asset prices
  1124. 40:43is a really hard problem. Uh this These
  1125. 40:46are incredibly noisy processes.
  1126. 40:49Uh very close to random walks.
  1127. 40:52Uh your out-of-sample R squareds, for
  1128. 40:54those of you who know what that that is,
  1129. 40:55are like low single digits.
  1130. 40:58So, like 0.03 is a successful R squared.
  1131. 41:02And, you know, anyone in any other sign
  1132. 41:04most other scientific fields would look
  1133. 41:06at that as a forecast out-of-sample R
  1134. 41:08squared and go,
  1135. 41:10"You're basically noise. Like you don't
  1136. 41:12have anything better than a a coin toss
  1137. 41:14here.
  1138. 41:15So, why are we doing this?"
  1139. 41:17But, that's a that's a successful That's
  1140. 41:20a successful outcome in my world, right?
  1141. 41:22So,
  1142. 41:24when you have that noisy of a process,
  1143. 41:26the techniques to
  1144. 41:28uh and by the way, very limited data.
  1145. 41:30Um
  1146. 41:31you know, if you're doing self-driving
  1147. 41:33cars, you can basically generate more
  1148. 41:34data just by So, like let's say they
  1149. 41:37have a self-driving car and, you know,
  1150. 41:39you find that it's not performing well
  1151. 41:41in rain.
  1152. 41:42Well, let's go to Portland and Seattle
  1153. 41:44for a while
  1154. 41:45and train more cars there. Like you
  1155. 41:47could just get more data.
  1156. 41:49For the capital markets, you can't get
  1157. 41:51more data. The data are non-stationary,
  1158. 41:54and there just isn't that much of it.
  1159. 41:56You know, if you compare like the number
  1160. 41:57of Facebook posts, the number of
  1161. 41:59Facebook users to like the total amount
  1162. 42:02of data about US equities.
  1163. 42:05And I mean, like the daily data in the
  1164. 42:07one versus the total cumulative data in
  1165. 42:10the other. It one like they're just
  1166. 42:12orders of magnitude different, you know?
  1167. 42:15Uh
  1168. 42:16we're talking about billions of data
  1169. 42:17points per day, if not more,
  1170. 42:19and on in the kind of social media
  1171. 42:22world, and we're talking about maybe
  1172. 42:26billions of data points in the whole
  1173. 42:27cumulative history of everything in US
  1174. 42:29equities.
  1175. 42:31And so,
  1176. 42:33you know, you can see that that's Those
  1177. 42:35are just different problems.
  1178. 42:37There are other differences, too, but
  1179. 42:38just for the sake of time, let's just
  1180. 42:40acknowledge that these are very
  1181. 42:41different problems.
  1182. 42:42You may need different and better
  1183. 42:44techniques and more powerful computing
  1184. 42:46clusters to be able to tackle these
  1185. 42:50differing challenges.
  1186. 42:51And so, that's the other aspect of the
  1187. 42:53supply side.
  1188. 42:55But again, the demand side is there.
  1189. 42:57So,
  1190. 43:00as conventional old-school techniques
  1191. 43:01have become
  1192. 43:03more heavily competed, as the
  1193. 43:05philosophies and concepts and ideas and
  1194. 43:09theories that under pin those strategies
  1195. 43:11have proliferated and been leaked out
  1196. 43:13into
  1197. 43:15competitive firms by virtue of employees
  1198. 43:17changing jobs or leaked out into the
  1199. 43:19academic journals and textbooks and so
  1200. 43:22on, you know, those old ideas are harder
  1201. 43:24and harder to make money on. And so,
  1202. 43:27just as some people are deciding, "Okay,
  1203. 43:29well, the hell with it. I'm going to go
  1204. 43:30venture far afield and and trade China,
  1205. 43:33even though it's a pain in the neck."
  1206. 43:35Uh
  1207. 43:36maybe
  1208. 43:37uh some of them are saying, "Well, I'm
  1209. 43:39going to venture further afield in my
  1210. 43:40techniques."
  1211. 43:42I happen to be really good at this
  1212. 43:43machine learning thing, and so I'm going
  1213. 43:45to do that, where it's harder to compete
  1214. 43:47with me.
  1215. 43:48So, there's that happening, too.
  1216. 43:50That there's a desire
  1217. 43:52to take
  1218. 43:54more fitting risk because the
  1219. 43:56lower-hanging fruit, which doesn't have
  1220. 43:57fitting risk, you know, it's not easy
  1221. 43:59making money that way anymore.
  1222. 44:02I feel as though a lot of what we've
  1223. 44:03been talking about here is very much
  1224. 44:06applicable to, you know, the large quant
  1225. 44:09funds, uh and very much the professional
  1226. 44:11space.
  1227. 44:12I'd like to ask you,
  1228. 44:14what have you seen in the trend among
  1229. 44:17DIY quants? So, the at-home retailer who
  1230. 44:22has, you know, might have some machine
  1231. 44:24learning skills, you know, has learned
  1232. 44:26how to program, has wrote some sort of
  1233. 44:28systematic strategies, cuz I think when
  1234. 44:31we last spoke in
  1235. 44:32uh 2016, you know, the past 4 years has
  1236. 44:35been a big rise
  1237. 44:37uh in this space.
  1238. 44:39Is there anything you've been seeing
  1239. 44:40here which is particularly interesting?
  1240. 44:45Yeah,
  1241. 44:46uh for sure. So, there's a few dynamics
  1242. 44:48there. It's a good question. And my
  1243. 44:50perspective on retail is mostly US and
  1244. 44:53then like random smatterings Asia. I
  1245. 44:56don't actually, weirdly, have any
  1246. 44:57impression at all of
  1247. 44:59Australia or Europe as far as retail
  1248. 45:01investing, but maybe you can tell me and
  1249. 45:03and we can reflect on it a little bit.
  1250. 45:04But, as far as
  1251. 45:07the US goes,
  1252. 45:08um retail investors were a steady kind
  1253. 45:12of part of that market
  1254. 45:14for a very long time. I can tell you
  1255. 45:16like when I was a child, I distinctly
  1256. 45:18remember my dad going through
  1257. 45:21the newspaper, like the physical print
  1258. 45:22newspaper,
  1259. 45:24and going through like which stocks he
  1260. 45:26owned and how they did yesterday. And
  1261. 45:27this is like the next morning finding
  1262. 45:29out how each stock he did. No clue how
  1263. 45:31his portfolio was like because you
  1264. 45:33didn't have to like sit there and do
  1265. 45:35that math.
  1266. 45:36But, you could certainly look at the
  1267. 45:38Wall Street Journal or the local paper,
  1268. 45:40and it would have a printout of
  1269. 45:42thousands and thousands of ticker in
  1270. 45:43very fine print,
  1271. 45:45and what their last close was, what the
  1272. 45:47change was yesterday, and what the
  1273. 45:48volume was, and a few other columns.
  1274. 45:51You know, folks like my dad have been
  1275. 45:53doing that for a long time.
  1276. 45:55Uh
  1277. 45:56when the internet kind of boom happened
  1278. 45:59in the late '90s,
  1279. 46:01and commensurate with that, tech stock
  1280. 46:04sort of IPOs boomed.
  1281. 46:06And, you know, pets.com and
  1282. 46:09and all that were out there. Um
  1283. 46:11So, now defunct company that sold like
  1284. 46:15pet care products online.
  1285. 46:17You know, there were a bunch of
  1286. 46:18companies that went public,
  1287. 46:20and
  1288. 46:22a lot of interest on those, but also
  1289. 46:23just in general on the markets because
  1290. 46:25the market was rallying really hard, and
  1291. 46:27it was pretty easy to kind of {quote}
  1292. 46:28{unquote} look like a genius. You bought
  1293. 46:30five internet companies and just were up
  1294. 46:32like 100% a week or whatever the hell.
  1295. 46:35And so, that dynamic really did lead to
  1296. 46:39a
  1297. 46:39pretty significant increase in the
  1298. 46:41participation of retail investors.
  1299. 46:43When that crashed,
  1300. 46:45that definitely took a a a big
  1301. 46:48chunk of retail participation with it
  1302. 46:50because those folks realized, "Well, you
  1303. 46:52know, it turns out this is hard, and I'm
  1304. 46:53kind of shitty at it."
  1305. 46:55And so, they stopped.
  1306. 46:57That took another leg down just a few
  1307. 47:00years later when the second that, you
  1308. 47:01know, 2008 crisis happened.
  1309. 47:04You know, the first crisis ended in
  1310. 47:052002,
  1311. 47:07late 2002, I think, maybe even early
  1312. 47:092003.
  1313. 47:11Uh and then the second crisis started in
  1314. 47:13November of 2007. So, you didn't even
  1315. 47:15have 5, 6 years between those two
  1316. 47:18crises. And that was two 50% drawdowns
  1317. 47:21in US equities in the space of 10 in the
  1318. 47:23space of 10 years. It's pretty crazy.
  1319. 47:26And unsurprisingly, that turned off a
  1320. 47:27lot of people
  1321. 47:29to investing in stocks at all.
  1322. 47:31And if they were going to invest in
  1323. 47:32stocks, it was just going to be passive.
  1324. 47:35So, that kicked off that passive trend.
  1325. 47:36It also kicked off a de-retailization
  1326. 47:39trend in the US.
  1327. 47:41Uh in some other markets, retail
  1328. 47:43investors have been more present uh for
  1329. 47:45longer and more consistently, like in
  1330. 47:48places like China and Singapore and so
  1331. 47:49on.
  1332. 47:50Um Korea.
  1333. 47:52Uh but, where there's And I don't know
  1334. 47:55how to say this without it sounding
  1335. 47:56politically incorrect, but like folks
  1336. 47:57just gamble. Like you go to the casino,
  1337. 47:59that's a completely normal thing to do.
  1338. 48:01In the US, that's not like a completely
  1339. 48:03normal thing to do. Casinos are like
  1340. 48:05illegal in most of the states in
  1341. 48:06America, which is a funny thing.
  1342. 48:09There's just less of that kind of
  1343. 48:11culture of just making bets and punting
  1344. 48:12around.
  1345. 48:13Um
  1346. 48:15which, for better or worse, is what it
  1347. 48:16is. Now, we've had this crazy bull
  1348. 48:19market,
  1349. 48:20and on top of that, a lot more tech is
  1350. 48:23available than has ever been before.
  1351. 48:25That's just the nature of things. That
  1352. 48:27same state statement will be true like 2
  1353. 48:29years from now, unless we
  1354. 48:31truly devour ourselves.
  1355. 48:34So, the platforms out there,
  1356. 48:36and I don't mean Robinhood, which is, to
  1357. 48:38me, just sort of a joke.
  1358. 48:40Uh I mean more
  1359. 48:42uh like Quantopian and other similar
  1360. 48:45platforms where anyone can get out there
  1361. 48:48and just
  1362. 48:49code up trading systems.
  1363. 48:52And then there's brokers like
  1364. 48:53Interactive Brokers, where if you're
  1365. 48:54like semi-pro, you have a few hundred
  1366. 48:56thousand dollars,
  1367. 48:58you can set up an account there and
  1368. 49:00have some of the same tools and
  1369. 49:04access as professionals have, although
  1370. 49:07it's
  1371. 49:08definitely a far cry, frankly, cost-wise
  1372. 49:10especially.
  1373. 49:12Yeah, it's, you know, it's kind of cool.
  1374. 49:14So, you can
  1375. 49:15you can set up like a one-person quant
  1376. 49:17shop, and I know lots of folks who do.
  1377. 49:19In fact, some of the folks we've
  1378. 49:20allocated to
  1379. 49:22come exactly from that lineage, where
  1380. 49:24they didn't go They weren't like
  1381. 49:26ex-employees of Two Sigma. Maybe they've
  1382. 49:28had some professional experience, but
  1383. 49:30what they really did was they're just
  1384. 49:32homegrown
  1385. 49:33hobbyists who are really good at it and
  1386. 49:35go, "You know what? We're going to We're
  1387. 49:37This is I'm good at this. I want to do
  1388. 49:38this professionally and scale up.
  1389. 49:40Uh and we've had some success with that
  1390. 49:42model. I mean, but you know, we've also
  1391. 49:44had some failure with that. It's not
  1392. 49:45easy.
  1393. 49:46But yeah, there's been this
  1394. 49:47kind of
  1395. 49:49uh combination or confluence of things
  1396. 49:52where we have this protracted bull
  1397. 49:53market
  1398. 49:55that's encouraged retail participation
  1399. 49:57and a lot of really specific brand
  1400. 49:59excitement around names like Apple and
  1401. 50:01Tesla.
  1402. 50:02And then at the same time
  1403. 50:05the availability of technology
  1404. 50:07to actually make it easy to code up
  1405. 50:10strategies and backtest them and
  1406. 50:12implement them and enter competitions.
  1407. 50:15Um and those things kind of combined to
  1408. 50:19to make it pretty
  1409. 50:21It's been Yeah, there's been like a big
  1410. 50:23surge in that world.
  1411. 50:25Yeah. I remember last time when we
  1412. 50:27spoke, you you mentioned you'd allocated
  1413. 50:30to someone who was essentially just a
  1414. 50:32one-man band working from his apartment.
  1415. 50:35Still have that money with that same
  1416. 50:36guy.
  1417. 50:37Yeah, I think you'd mentioned he was
  1418. 50:39managing about $15 million. I'm not sure
  1419. 50:41what that is today, but 70.
  1420. 50:4470 million, one guy from his apartment.
  1421. 50:47Mhm. That's incredible. That's not 70
  1422. 50:50million from you. That's you're just a
  1423. 50:52portion of that. No, it's all from us.
  1424. 50:53All from you. Yeah. Wow.
  1425. 50:58That's that's uh quite incredible. Let
  1426. 51:00me ask you this question
  1427. 51:02and we'll wrap things up here. But let's
  1428. 51:04say there is someone listening to this
  1429. 51:05who kind of fits into that category of
  1430. 51:08DIY quant,
  1431. 51:10you know, retailer working at home.
  1432. 51:13They've developed a quantitative
  1433. 51:14strategy. They've been having a little
  1434. 51:15bit of success with it, trading it for
  1435. 51:17some time, real money.
  1436. 51:20They like the idea of
  1437. 51:21running more capital through that
  1438. 51:23strategy, scaling up. What would this
  1439. 51:26trader need to do to be ready for
  1440. 51:28investment?
  1441. 51:30That's a great great question, Aaron.
  1442. 51:33So,
  1443. 51:34it depends is the answer.
  1444. 51:37Um if
  1445. 51:38they are willing to simply license their
  1446. 51:40signal, meaning
  1447. 51:42they would once a day or on whatever
  1448. 51:44rate the right frequency is for that
  1449. 51:46strategy, send a list of trades that
  1450. 51:49their model comes up with
  1451. 51:51to someone who has all the
  1452. 51:52infrastructure and compliance and
  1453. 51:54operational aspects to do that.
  1454. 51:56Um to to put on those trades and do it
  1455. 51:59correctly.
  1456. 52:00Uh then the bar is a lot lower because
  1457. 52:03they don't really need much of anything.
  1458. 52:06If they want to actually like set up a
  1459. 52:09fund
  1460. 52:10and or a firm that does this and, you
  1461. 52:13know, you'll have to register with the
  1462. 52:14relevant authorities as a as an
  1463. 52:17investment advisor or whatever the
  1464. 52:18equivalent is in your jurisdiction. Um
  1465. 52:21you'll have to show to investors that
  1466. 52:23you can actually process trades and all
  1467. 52:26kinds of other things. Um
  1468. 52:29You know, that hurdle's a lot higher,
  1469. 52:30like a lot higher. And it's expensive
  1470. 52:33and not likely to work. So, just to be
  1471. 52:35clear.
  1472. 52:36Um you know, I think that the
  1473. 52:39far more sensible path is to contract
  1474. 52:43with someone who's already got a
  1475. 52:45substantial operation
  1476. 52:47and
  1477. 52:49license your signals out.
  1478. 52:51And then if it turns out you're really
  1479. 52:52good at this in the long run, A, you'll
  1480. 52:54just make money doing it. You may not
  1481. 52:55have to quit your day job.
  1482. 52:58But B, if you decide you really want to
  1483. 52:59do it, at least you'll have some track
  1484. 53:01record
  1485. 53:02of what your signals have done. And then
  1486. 53:05when you try to go out and get unreal
  1487. 53:06size, you know, and then it at least
  1488. 53:09when you try to go out and get capital
  1489. 53:11one of the big questions that investors
  1490. 53:13ask, which is is this thing scalable?
  1491. 53:16You know, if I run a $500 billion
  1492. 53:18pension fund or a $10 million fund of
  1493. 53:21hedge funds or family office or
  1494. 53:23whatever.
  1495. 53:24Um I'm not looking to make
  1496. 53:26a $10 million investment in a hedge fund
  1497. 53:29that might return 10% per year. Like
  1498. 53:31making a million dollars of P&L per year
  1499. 53:34for someone at that size is like not
  1500. 53:36worth the effort.
  1501. 53:39Not worth the marginal cost. And so, you
  1502. 53:41got to really demonstrate that you are
  1503. 53:44uh able to scale. And if you can
  1504. 53:46demonstrate that, oh no, my signal
  1505. 53:48generated 20 million of P&L
  1506. 53:50admittedly it wasn't my firm that that
  1507. 53:52put the trades on.
  1508. 53:54Then yeah, there's there's something
  1509. 53:56there. But the more realistic thing is
  1510. 53:58that if you're a hobbyist, stay a
  1511. 53:59hobbyist, frankly, and license your
  1512. 54:01signals to people who
  1513. 54:03are doing this for real and
  1514. 54:06um
  1515. 54:07and you get paid.
  1516. 54:08Uh and that's kind of cool. What would
  1517. 54:10be an example of someone who may be
  1518. 54:12interested in
  1519. 54:13in licensing signals though? Like would
  1520. 54:16that be
  1521. 54:17a fund manager or
  1522. 54:19Yeah, so um and I I think they've
  1523. 54:21stopped doing this now, but WorldQuant
  1524. 54:24used to do it and globally. Like they
  1525. 54:25had
  1526. 54:27farmers in Vietnam who were contributors
  1527. 54:29of signals. Right. Um but I think
  1528. 54:32they've shut that practice down. I'm not
  1529. 54:34100% sure.
  1530. 54:36Yeah, we know there's there's there's a
  1531. 54:38few out there. I mean
  1532. 54:39some of them are looking for more like
  1533. 54:41professional CVs, like people who used
  1534. 54:44to work
  1535. 54:45at at firms that are doing this
  1536. 54:47professionally and maybe have
  1537. 54:50gone off on their own.
  1538. 54:52Yeah, there there are a few out there.
  1539. 54:54I'm not comfortable naming names to be
  1540. 54:56honest just because No, I wouldn't
  1541. 54:58expect you to. Yeah, I don't have But
  1542. 55:00yeah, these are professional quant hedge
  1543. 55:01funds. Okay.
  1544. 55:03Would you have any luck, you know, maybe
  1545. 55:05with some smaller family offices? Do you
  1546. 55:07think that's something they may be
  1547. 55:08interested in?
  1548. 55:10Sure, if you have a relationship. But
  1549. 55:12you know, as with so many things, like
  1550. 55:14these days there's so many people
  1551. 55:15unemployed, especially in the US.
  1552. 55:17That
  1553. 55:18their resume is going into the inboxes
  1554. 55:20of every person that's every employer
  1555. 55:22that's got a job opening.
  1556. 55:23It's not quite hopeless
  1557. 55:26if you randomly send your resume into a
  1558. 55:28vacuum, but you definitely know it's a
  1559. 55:30lot better if you have a connection.
  1560. 55:31Well, how much more so for something
  1561. 55:33like this where you're asking for
  1562. 55:34millions of dollars or tens of millions
  1563. 55:36of dollars
  1564. 55:37of risk capital and a budget.
  1565. 55:40Um that you know, that's not just a
  1566. 55:41normal job, right? That's a job with
  1567. 55:44some real strings attached. So, if you
  1568. 55:45have like a good relationship with
  1569. 55:46someone, then sure, it's I think worth
  1570. 55:48barking up that tree, but
  1571. 55:50I I honestly think it's
  1572. 55:52like lower probability than a lottery
  1573. 55:54ticket if you're just like casting your
  1574. 55:57net out into the ether.
  1575. 55:58Trust is a big part of this, you know?
  1576. 56:01For sure. Yeah.
  1577. 56:03I know you've got to run. I was going to
  1578. 56:05ask you really quickly
  1579. 56:08about your involvement with DARPA. Is
  1580. 56:09that something you'd be interested in
  1581. 56:11talking about?
  1582. 56:13Yeah, but I'm not like under NDA or
  1583. 56:14anything and I can talk about it really
  1584. 56:16briefly.
  1585. 56:17Um
  1586. 56:18It So, it was really cool and fun and
  1587. 56:20interesting. So, basically uh
  1588. 56:22some years ago I was approached by
  1589. 56:25um a professor who's a friend of mine
  1590. 56:28who is friends with some folks who are
  1591. 56:30at DARPA
  1592. 56:31or were at the time and said, "Hey,
  1593. 56:33would you be interested in connecting
  1594. 56:34with
  1595. 56:35They were looking at uh financial
  1596. 56:37warfare kind of things.
  1597. 56:40And so, this is part of the Department
  1598. 56:41of Defense, the US Department of
  1599. 56:42Defense. And this was not an exercise in
  1600. 56:46looking at, you know, aggressive
  1601. 56:48tactics, rather in defensive tactics and
  1602. 56:51strategies for
  1603. 56:54protecting American markets and
  1604. 56:57financial systems from
  1605. 57:00various parties that for either
  1606. 57:01political or economic reasons
  1607. 57:04may want to
  1608. 57:07cause problems, right? So, causing
  1609. 57:09outages at exchanges or causing um
  1610. 57:13outages with, you know, payroll
  1611. 57:15processing systems or
  1612. 57:17any number of other vectors of threat.
  1613. 57:19And there are of course continuously
  1614. 57:21both governmental and non-governmental
  1615. 57:23parties that are very aggressive in
  1616. 57:27in that stuff. And so, it's, you know,
  1617. 57:28it was this super fun intellectual
  1618. 57:31exercise of like what are the vectors of
  1619. 57:33threats
  1620. 57:35detect them early, prevent them
  1621. 57:37mitigate the spread of the damage.
  1622. 57:39Um so, you know, we do things like war
  1623. 57:41games and
  1624. 57:42um
  1625. 57:44and table exercises to to sort of flesh
  1626. 57:47out some of these things a little bit.
  1627. 57:49And the part of the goal was to see is
  1628. 57:51there something
  1629. 57:53measurable and quantifiable and within
  1630. 57:54DARPA's remit to actually focus on
  1631. 57:58to uh
  1632. 58:00to work on this problem. So, the way
  1633. 58:01DARPA works is it starts as a seedling,
  1634. 58:03which is all this thing has been so far.
  1635. 58:05And then it only becomes a program
  1636. 58:08if the seedling shows that there is
  1637. 58:11not only something to do here, but
  1638. 58:12something for DARPA specifically to do
  1639. 58:15because DARPA isn't, you know
  1640. 58:16doesn't have a completely
  1641. 58:18carte blanche remit. It has a specific
  1642. 58:20remit. And so, if it if there's stuff to
  1643. 58:23do, but it's better for like the
  1644. 58:24Treasury Department to do, well, then
  1645. 58:26that's fine. Like hand that off to the
  1646. 58:28Treasury Department. They can decide
  1647. 58:29what to do.
  1648. 58:30You know, if it's it's for DARPA, then
  1649. 58:32it has to fit some criteria. And
  1650. 58:35uh so yeah, they're still in the
  1651. 58:37evaluation phase. It does appear that
  1652. 58:40there's enough for some stuff to be done
  1653. 58:42as a program. But that has not been
  1654. 58:45finalized.
  1655. 58:47It's super fun, really interesting. It
  1656. 58:50does sound like an interesting project
  1657. 58:51to be involved with. So, you're just
  1658. 58:53sort of consulting on it at the for the
  1659. 58:55time being?
  1660. 58:57Yeah, and I mean I'm I'm like a
  1661. 58:58volunteer. Uh so, I'm consulting
  1662. 59:01informally, no security clearances, no
  1663. 59:03NDAs.
  1664. 59:05I literally had one condition, which is
  1665. 59:06that I get to put it on my LinkedIn cuz
  1666. 59:09how cool is that? Yeah, yeah.
  1667. 59:11Exactly. That's why I asked you about
  1668. 59:12it.
  1669. 59:13Yeah.
  1670. 59:14And I I wasn't really sure the link
  1671. 59:16between DARPA cuz I just know them as
  1672. 59:18creating the the robot dog.
  1673. 59:21And other crazy robots. I wasn't really
  1674. 59:22sure the link between that and financial
  1675. 59:24markets, but but yeah.
  1676. 59:26It was a little bit surprising, but you
  1677. 59:28can see that it's
  1678. 59:29you know, economic defense and military
  1679. 59:31defense are sort of more interlinked
  1680. 59:33than ever, you know. Yeah. Yeah.
  1681. 59:34Many wars have been fought over economic
  1682. 59:36interests, I think far more than people
  1683. 59:38generally realize. Right.
  1684. 59:41For sure. Well, Rishi, I know you need
  1685. 59:43to shoot off. Uh I just want to say uh
  1686. 59:45it was really nice to catch up with you
  1687. 59:47and uh thank you very much for coming on
  1688. 59:49the podcast and um share a bit of an
  1689. 59:51update on some of the things you've been
  1690. 59:52seeing and what's been developing in
  1691. 59:55markets. So, if someone wants to follow
  1692. 59:58you, uh would you like to share your
  1693. 1:00:00Twitter handle and maybe your website
  1694. 1:00:02and also mention your book, too?
  1695. 1:00:04Sure, Aaron. Thanks a lot for having me
  1696. 1:00:06again. Again, good to catch up. Uh
  1697. 1:00:09I'm on Twitter, not super actively, but
  1698. 1:00:11I am on Twitter
  1699. 1:00:13uh at Rishi K. Narang.
  1700. 1:00:15And
  1701. 1:00:18and I did write a book, the second
  1702. 1:00:20edition of which came out in 2013,
  1703. 1:00:23called Inside the Black Box. That should
  1704. 1:00:25be readily available on places like
  1705. 1:00:27Amazon.
  1706. 1:00:28Uh what was the other thing I was
  1707. 1:00:30supposed to share?
  1708. 1:00:31Uh your website.
  1709. 1:00:33I don't really have one. Okay.
  1710. 1:00:36The T2 website, no?
  1711. 1:00:38Uh yeah, there's nothing on there. But
  1712. 1:00:40yeah, my company is www.t2am.com.
  1713. 1:00:44Okay. And Inside the Black Box is a
  1714. 1:00:47fantastic book. I've read it a
  1715. 1:00:49several times. Uh I have a copy myself.
  1716. 1:00:51Uh would definitely recommend it if
  1717. 1:00:52you're interested in this type of thing.
  1718. 1:00:54So, yeah, until next time, Rishi.
  1719. 1:00:57Uh we'll we'll chat soon.
  1720. 1:00:59Thanks, Aaron. Stay safe, man. Be good.
  1721. 1:01:03You've reached the end of this episode
  1722. 1:01:05of Chat with Traders. But rest assured,
  1723. 1:01:07there are more episodes loaded with real
  1724. 1:01:10market insight and zero hype on the way
  1725. 1:01:13soon. So, to stay updated with each
  1726. 1:01:15great new release, subscribe to the
  1727. 1:01:17podcast on iTunes. And we'd love it if
  1728. 1:01:19you'd leave a rating and review. We'll
  1729. 1:01:21catch you next time on Chat with
  1730. 1:01:23Traders.

About this transcript

This page contains the full transcript of Evolving Trends: How Quants Are Earning Alpha · Rishi Narang by Chat With Traders, generated from the public captions YouTube serves with the video. The transcript has 9,954 words across 1,730 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.