YT YouTube2Text

YouTube2Text → Transcripts → My Python strategy with 100% win-rate (+ live results)

My Python strategy with 100% win-rate (+ live results) — Transcript

by Algo-trading with Saleh · 4,415 words · 606 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00Today's strategy has a 100% win rate.
0:02Now, I know what you're thinking. That
0:04sounds crazy, right? Well, stick with me
0:06because first I'm going to explain to
0:08you how the basic math behind it works.
0:10Then I will write the code in Python so
0:12we can run some back test on it. And
0:13then I will even show you some of my
0:15real trading results with this strategy
0:17from years ago. And we will discuss
0:18whether or not you should trade the
0:20strategy in the first place. So, if that
0:22all sounds good, let's get right into
0:23it. Now, this strategy is called Marting
0:26Gale and it was originally a gambling
0:28technique. So needless to say it is
0:30extremely risky but if implemented
0:33correctly it works at least in theory.
0:35So let me explain the math behind it
0:37with a very simple example. So let's
0:39imagine that the cost of one BTC is $100
0:42and we decide to open a long position at
0:45this price. That means the size of our
0:46position is going to be one BTC and the
0:48amount that we have spent so far is also
0:51100 bucks. Nothing complicated. If the
0:53price goes in our favor, let's say it
0:55reaches
0:56$110, then we're going to make $10 per
0:59each BTC, which in this point we only
1:01have one, that means we've made $10 in
1:04total profit. Now, of course, this is
1:05assuming that we sell the position here.
1:07But what if the price actually goes
1:09against us, for example, it reaches $80.
1:12So, this means at this point, we are 20
1:14bucks in a loss if we close the
1:15position, of course. But what if we
1:17decide to actually double the size of
1:19our position? So, not only we don't
1:20sell, we even double down. literally. So
1:23now the new entry price is going to be
1:25100 + 80 divided by two which means 90
1:28bucks. So again the price is here but
1:31our entry is no longer at 100 bucks. It
1:34is now at 90. Now what could happen at
1:36this point is the price could continue
1:38to go down or it may go up. For example,
1:40if it reaches 90 bucks, we're going to
1:42break even. Now of course we also need
1:44to consider trading fees. But as a
1:46matter of fact, years ago when I was
1:47trading the strategy, I was using this
1:49exchange called BitMX, which isn't as
1:51popular these days. Now, back then, they
1:53were offering trading fee rebates for
1:56market makers. And because this strategy
1:58uses limit orders, that means I was
2:00getting trading fee rebates. So, not
2:02only I wasn't paying the exchange for
2:04trading, I was getting paid for it. That
2:06means even if I close my position at the
2:08same price that I opened it, I would
2:10have still made money. All right? But
2:12let's imagine that the price continues
2:13to go up and it reaches $100 again.
2:16Well, at this point, we're going to be
2:18in profit. Exactly how much? Well, it's
2:21going to be $10 per BTC, but because we
2:24have two of them, it's going to be 20
2:26bucks in total profit. So, you see, even
2:28though we reached the current price that
2:29we opened this trade in the first place,
2:31we are now in a profit. Again, that's
2:33because our entry price is now at 90
2:35bucks. But let's talk about what if the
2:37price continued to go down. So instead
2:39of 80 bucks, it was now at 60. And we
2:42decided to double down again. So now we
2:44have two BTC, but we purchase again at
2:46the new price of 60. Again, the same
2:49size of the position that we had. So we
2:50had two BTC, but to double down, we're
2:53going to purchase two more. That so that
2:55means now we're going to have four BTC.
2:57And the total amount of money that we've
2:59spent on this trade so far is going to
3:01be $300. And the new entry price is
3:03going to be 300 divided by 4, which
3:05equals 75. So this is our new entry
3:09price. So now if the price goes up,
3:11let's say it reaches 85, we're going to
3:13be at 40 bucks in total profit. Again,
3:16$10 per BTC, but because we have four
3:18BTC, now we are sitting in this price.
3:21So you see, even though we started this
3:22trade at the price of $100, because it
3:25had some price fluctuation, now even if
3:27the price reaches $85, we're going to be
3:30in a profit. So we didn't even have to
3:33wait for the price to go above 100. So
3:36that's really the beauty of this
3:37strategy. You don't even need the price
3:39to go in your favor. In fact, the more
3:41it goes against you, the more money
3:43you're going to make. Now, of course,
3:45this assumes that it does go in your
3:47favor, at least a little bit at one
3:49point, but I'll get to that in a minute.
3:50So, now let's open Jess's dashboard and
3:53create a new strategy. I'm going to call
3:55this Martin
3:58girl. Open my editor and look it up. All
4:01right. So, first we need to define our
4:03interview rule. And for that I'm going
4:04to make it super simple. I will simply
4:06return true. That means as soon as I run
4:09the strategy I'm going to try to open
4:11the long position. And it is
4:12significantly easier to write the long
4:14version of the strategy. So I'm going to
4:16remove the short functions
4:18entirely. So I only have these two now.
4:22Now to do the position sizing and submit
4:24the actual order. First we need our
4:26entry price. Now I'm going to say entry
4:28equals I want the current Ballinger band
4:31value. So let's define a new
4:35property and I'm going to call it BB
4:38short for Ballinger
4:40bands and in it we will simply return
4:43TA. Ballinger bands and then I will pass
4:47the current candles. Now in case you're
4:48not familiar with Ballinger bands, let's
4:50look it up on Trading
4:53View. It's basically very simple
4:56indicator and it has one moving average
4:58which is this blue line. And then we
4:59have the lower band and the higher band
5:02which are simply two times of the
5:03current SD or the standard deviation.
5:06Now this indicator is great for ranging
5:08markets. You could simply buy when the
5:10price reaches the lower band and sell
5:13when it reaches the mean which in this
5:15case is this moving average, this blue
5:17line. You can also open a short position
5:18when the price reaches the upper band
5:20and then again sell it when it reaches
5:22the mid. But in reality it's not that
5:25easy to write a profitable strategy with
5:27it. Now all I care about is the lower
5:29band because let's say at this point I
5:31decide to open a long position. I want
5:34my entry order to be a limit buy order
5:36at the lower band. So again for example
5:39let's say we decide here then the buy
5:42order is going to be here. So let's go
5:44here. I'm going to say
5:46selfb lower band is going to be our
5:49entry price. All right. Next I need my
5:51quantity and I'm going to use utils size
5:55to quantity. first parameter I will pass
5:57the current available margin but I don't
6:00want the entire amount for my first
6:02order because remember we're going to
6:03submit multiple orders whenever the
6:05price goes against us right so I'm going
6:08to begin with 10% of that next I will
6:11pass the current entry order and as fee
6:14rate I will pass the current fee of the
6:16exchange that I'm trading next I will
6:18simply submit my buy order by saying
6:21self buy equals quantity and then the
6:24entry And that's it. All right. So now
6:27we have the entry of the strategy. But
6:29what about exiting it? But for that
6:31first I'm going to have to define the
6:33unopen position which will get executed
6:37when the position is open which is the
6:39moment we want to submit our take profit
6:41order. Right? So let's scroll this down
6:43a bit. All right. So I'm going to say
6:47the TP price standing for take profit
6:51price
6:52equals the current positions entry price
6:57plus some kind of value. Now again
7:00remember this diagram I showed you. So
7:02if this is our entry price, we want our
7:04exiting price with a profit to be
7:06slightly above it like for example in
7:08here. But what should we use for that
7:10number? Like in this example, we simply
7:12use the number 10. But in an actual
7:14strategy, we need a dynamic value. And I
7:17think one of the best ways to do that is
7:19to use either the ATR indicator or the
7:22standard deviation. So I'm going to say
7:23self std multiplied by two. But we
7:27haven't defined this std yet. So let's
7:30go up, come down, and define a new
7:34property. In it, we simply return TA.
7:39STDEV, which is stands for standard
7:41deviation. And then we pass the current
7:43candles and this should give us the
7:46current SD value. Now, by the way, in
7:48Python, you don't have to specify the
7:50return type of the function, but if you
7:52do, it would be nice and help your
7:54editor a little bit. But anyways, let's
7:56go back. So now this value actually
7:59works. Next, I need to submit the actual
8:01takeprofit order. For that, I'm going to
8:03say
8:04self.takerprofit equals. And as the
8:06quantity of this order, we will simply
8:08use the quantity of our current
8:10position. So I will say self position
8:13dotquantity and as the price I will pass
8:15TP price and that's it. Now we have our
8:17take profit. But if you remember this
8:19example once we open the position we
8:22also need to submit our next entry order
8:24or in other words the order which if the
8:27price reaches it is going to double the
8:29size of our current position. So let's
8:31actually add a comment here saying take
8:35profit and next I will submit our next
8:40entry. All right. So I'm going to say
8:42next
8:43entry price is going to be the current
8:48price
8:50minus again the current
8:53SD multiplied by two. And the next
8:58entries quantity is going to be the size
9:01of our current position because again
9:03that order is supposed to double the
9:05size of our position. So if we buy as
9:07much as we already had that's going to
9:08double the size. So that's going to be
9:10simply the current position's quantity.
9:13All right. Now to submit the next entry
9:15order. Remember initially we use self
9:18buy to also add to the size of it. It's
9:21going to be another buy order, right?
9:23And it is an entry order. So it's not
9:25the current takeprofit order or the
9:27current stop loss. So I will simply
9:30say self buy equals next entry quantity
9:36and then next entry price and that's it.
9:39Now remember that at the point where we
9:42are executing this code, we already have
9:45an open position. So that means the
9:47previous self.Y variable that we filled
9:50in here is no longer needed. it has
9:53already been executed and emptied behind
9:55the scenes for us. So don't worry about
9:57it. All right. So now we are submitting
9:59the take profit order and our next entry
10:02order as soon as the position is open.
10:05But if you remember our simple example,
10:06we may have to enter the position or
10:09let's say add to the size of our
10:10position multiple times and not just
10:12once. So doing this just once when we
10:16open the position is not enough. So we
10:18need to tell Jesse to submit the next
10:20entry order and also update our
10:23takeprofit order whenever we increase
10:25the size of the position. So for that
10:26I'm going to use this function called on
10:29increase position. Now in here we will
10:32simply do what we did in here or in
10:35other words let's refactor this. So I'm
10:37going to write another function and I
10:39will simply call it something like
10:42submit exits
10:45orders and paste this
10:48here. And in here I will remove this and
10:53simply
10:57call this. And one more time I will do
11:00it in here. So now not only we submit
11:04our take profit order and the next entry
11:07order when the position is opened, we
11:10also do the same thing when the position
11:11size is increased. All right, there's
11:13one more thing left because we are using
11:16a limit order to enter the position. I
11:18want to make sure that we keep updating
11:20that order. So let's say we decide to
11:22enter the position here and we submit
11:24the entry order which again is a buy
11:26limit order. We submitted at this price,
11:29but the price doesn't reach it. In the
11:31next candle, let's say the price still
11:33didn't reach it. But if you notice, the
11:36lower band is going down. So, I want to
11:38make sure the entry order is getting
11:39updated. So, when the price hits it, it
11:42is sitting exactly at the current
11:44Ballinger band price. To do that, we
11:46need to tell Jesse that if this order
11:48here wasn't executed yet and another
11:51candle has already closed, I need you to
11:54cancel the previous order and submit
11:57another one. Now to do that, we should
11:59simply say def should cancel in tree
12:02order and in it I will simply return
12:05true because I want to cancel all the
12:07time. Like if it hasn't already been
12:09executed, I want to cancel it because I
12:11know if it gets cancelled, it will come
12:13here because by doing so, it will
12:15re-execute this block of code. It will
12:17recalculate everything and submit the
12:19next order with the new price. All
12:21right, that's it. But there's just one
12:23more change I want to make in the std
12:26here. Let's rename this to make it a
12:30little bit simpler. So I'm going to call
12:32it margin because it's that distance
12:35between the current price and the next
12:37price or in other word the margin the
12:40difference. So I will call it this
12:42because I don't want it to just be the
12:44current standard deviation. I want it to
12:46be that multiplied by two. Now you could
12:48also play around with this code. You
12:50could increase it or lower it or even
12:52use something else such as the current
12:53ATR. Now remember that when I renamed
12:56this, it also changed it wherever we
12:58were using it, such as this and this.
13:00All right, now that we have everything,
13:02we're ready to actually back test this.
13:04All right, so let's go to Jess's
13:05dashboard and look up the word marketing
13:08gap, which is the strategy we just
13:09wrote. The exchange is Binance Perpetual
13:11Futures. The symbol is soul UCT. The
13:15time frame is hourly and the duration of
13:18it is the first four months of 2025. Now
13:22I have also turned on the interactive
13:25chart. Let's turn off the debug mode. I
13:28don't need it right now. The fast mode
13:30is on and so is the benchmark feature.
13:33Now one important thing about the
13:35strategy is that we're going to need a
13:37lot of capital. Like the more the
13:39better. And the maximum leverage on most
13:42exchanges is 100. So because of that I
13:44have made sure that my leverage number
13:46is also 100. Now let's run it.
13:52And here's the result. So the net profit
13:56is
13:58186%. We paid this much in trading fees.
14:01The draw down is minus
14:0443%. The win rate is at 100%. And the
14:09sharp ratio is at 2.66. All right. So
14:13let's also take a look at the
14:14interactive charts. So here you can see
14:17the trades that it executed. So we
14:20bought here and we sold here in a
14:21profit. We bought here, we sold in a
14:23profit. We bought, we added again, added
14:26again, added again, added again, and
14:28then again. And then again two more
14:30times and then sold everything here for
14:34a nice juicy profit. So by the way, you
14:36can see the quantity of the trades also
14:39in here. Then again we bought
14:42here and
14:44here and here
14:46again until we close the position in
14:50here for a nice juicy profit. And if I
14:52browse here you can see that all of them
14:54are wins. And if I click on one of them
14:56you can also see the details of it. So
14:58for example the entry price the exit
15:00price the entire quantity of the trade
15:03the P&L how many percentage it was and
15:06things like that. And notice that all
15:09the trades are limit orders. All right,
15:11so let's close this one
15:14and do it again with the two hours time
15:17frame. So while that's going, I want to
15:19quickly remind you guys about Apex Omni,
15:21which is my favorite decks made by the
15:23team behind Bybit. It has low trading
15:26fees, instant order execution, and of
15:28course, no KYC whatsoever. If you're
15:30going to check them out, please consider
15:32using our link, which not only will give
15:34you trading fee discounts, but it will
15:35also support me in order to make more
15:37videos like this. Again, we are in a
15:40profit. The draw down is now lower. The
15:44M rate is still at 100%. The sharp is
15:47now at
15:492.54. Let's do it one more time for the
15:525 minutes time frame. Let's look up that
15:56strategy.
15:57But notice that I have changed the
15:59duration now. So if I run it with the
16:02five minutes time frame, here's the
16:05result. We executed 19 trades over 9
16:09days. We made more than 5% in profit.
16:13The win rate is again at 100%. The sharp
16:16is at 7.74, which is absolutely
16:21crazy. Now, like I said, you can run
16:24this strategy on any other symbol. So
16:26let's try it on
16:29ETH on the hourly time
16:33frame over the duration of 1 month. Here
16:36is the result. We made 90% in
16:40profit. The draw down is minus 10%, the
16:43M rate is again at 100%. And we can also
16:47look at the trades. So in a range it
16:49absolutely crushed it.
16:52And in a downtrend, you might think this
16:56was a bad trade, but because the size of
16:58it increased, when we closed it
17:00eventually here, it actually made the
17:03most amount of money. So, this is the
17:07trade. It made us this much in profit.
17:10All right. So, the strategy seems
17:12perfect, right? So, at this point,
17:14you're probably thinking to yourself, if
17:16the strategy is literally perfect, like
17:19why is he even making this video, right?
17:21like why isn't he a billionaire by now?
17:23Well, let me show you my actual trading
17:25results with this strategy. So, this is
17:28for years ago for 2018 on the exchange
17:32BitMX as I explained earlier and you can
17:35see the sizes of the trade. So, for
17:37example, I shorted here then I closed
17:40the position here. I shorted, shorted
17:42and then closed it. Went long and then
17:45closed it. Shorted, shorted and then
17:47closed it. So again every time we start
17:49with some number then we keep doubling
17:52the size of the position which means our
17:54closing order is also the size of the
17:56entire position. Now again shorted
17:58shorted shorted shorted and I kept
18:00doubling and then I closed it. And the
18:02more I did this actually the profit was
18:05bigger. You can also see the trading fee
18:07we base that I was talking about. It was
18:10this number. You see it's negative. It
18:12means the exchange is actually paying
18:14us. We are not paying the exchange. And
18:16you can also see that these orders were
18:18submitted via the API. And the asset I
18:21was trading was called XBT, which is the
18:23name of BTC on
18:25BitMX. This again is some other trades.
18:28So it wasn't just one or two lucky
18:31trades. This went on for 48 hours. Yes,
18:34you heard that right. 48 hours. Because
18:36after 48 hours, what happened was this.
18:41Look at the last trade. It is
18:44liquidation. And yes, I got liquidated
18:46and the whole account was gone. And
18:48that's really the issue with this
18:50strategy. And it's a big one because
18:52you're going to win all the time until
18:55you don't. And when you don't, you don't
18:58just lose a little bit of money. You get
19:00liquidated. Game over. Now, let me
19:02explain why this is happening. So,
19:04imagine we have 500 bucks for our
19:07trading capital. If we have 100
19:09leverage, that means in total we have
19:11$50,000. Then assuming we have 100x
19:14leverage, we're going to have $50,000
19:16purchasing power. Now, here's why this
19:18is important. Because let's say the
19:20price of one BTC is 500 bucks and we buy
19:23one BTC and for that we're going to
19:26spend $500, right? And the size of our
19:29position is going to be one BTC. In the
19:31next step, when the price goes against
19:33us, and to keep this example easy, I'm
19:35going to assume that the price keeps
19:37going 20 bucks against us. All right? So
19:40the next step, we're going to have to
19:42submit another order again with the size
19:45of one BTC to double the size of our
19:47position. And the value of this new
19:49order is going to be 480 bucks because
19:52now that is the current price. The
19:54entire amount that we have spent so far
19:56is going to be 980 bucks and the size of
20:00the position is going to be 2 BTC. The
20:02next step, let's say the price is now at
20:04460. The size of the order is going to
20:07be two and the value of it will be this
20:09much. And now the entire capital we have
20:12spent so far is going to be
20:15$1,900. And again the size of the
20:17position is at four. So we will keep
20:20doing this. And notice that the entire
20:22amount of capital that we need is keep
20:24getting almost doubled. And it's not
20:26like we have infinite amount of money,
20:28right? So how much can we take this?
20:30Well, according to my calculation, we
20:32are only able to keep doing this until
20:35seven steps because at the seventh step,
20:38the amount of capital we have spent so
20:40far is
20:43$48,620. And again, this assumes our
20:46purchasing power was $50,000. So that
20:48means we are not able to make this 8
20:51accept which requires us to have
20:54$92,000 in capital. If we did, we would
20:57have been able to continue of course. So
20:59that's really the issue with this
21:00strategy. Yes, it seems perfect, but
21:03only if you have infinite amount of
21:05money, which we obviously don't. So now
21:07let me show you this. So let's say for
21:09the one with the time frame of 5
21:10minutes, what would have happened if I
21:13increase the ending date to the next
21:17month, we're going to get an error like
21:19this. It's telling me you cannot submit
21:21an order with a value of this amount
21:24when your available margin is this much.
21:26Now it's telling me consider increasing
21:27the leverage which I already have. I set
21:30it to 100, right? So I cannot do that
21:32anymore. Basically what's happening is
21:35that it's trying to submit the next
21:37order but it's running out of capital
21:39because this is a bit different than
21:41getting liquidated because it's trying
21:42to submit that last entry order but it's
21:45running out of capital. Which means if
21:47in reality in live trading you weren't
21:49able to submit this order, which would
21:51have been the case, you would have had
21:53to hold the position and the price would
21:55have gone against you until the point
21:57where you would have gotten liquidated.
22:00There you go. That's the issue with the
22:02strategy. Now, you might be thinking,
22:04what if we used a stop loss? What if we
22:06didn't keep the position or kept adding
22:08to it until this point, right? What if
22:10we accepted our losses, exited the
22:13trade, and started another one? Well,
22:16that is an option. You're welcome to try
22:18it, but I highly discourage you to do
22:21that because if you want to be a
22:22successful trader in the long term, you
22:25need to have a solid risk management in
22:27place. And for that, you want to exit
22:29the trade as soon as it is going against
22:31you. You want to keep your losses short
22:33and let your wins to be big. But in a
22:36strategy like this, you actually make
22:38the most amount of money when it is
22:40going against you. So if we had exited
22:43the trade early, we would not have made
22:45these numbers. So we had to let it to go
22:47against us. And that's really against
22:50all the values that I have for risk
22:51management in trading. So if you don't
22:53use a Starbucks like this example,
22:55definitely don't do it because in live
22:57trading, you're going to face all sorts
22:58of issues. Your internet connection
23:00might go down, the exchange might go
23:02down, or even maybe the software you are
23:04trading it with might have some bugs.
23:06All these things could and will happen
23:08at some point. And you want to make sure
23:10that if something went wrong, you only
23:12lose one trade, not your entire trading
23:15account. So because of that, you
23:16absolutely shouldn't trade this strategy
23:18without a stop. And if you do use a
23:20subplus, you're not going to get good
23:22numbers. But nonetheless, I think this
23:24was a good type of a mean diversion
23:26strategy. And there are some good ideas
23:29about it. And we might be able to take
23:31this concept, work on it to develop a
23:33good means strategy in the future. So, I
23:36hope you guys enjoyed this video and
23:37maybe learned at least one or two things
23:39from it. Now, we're going to have a
23:40giveaway. A random person who likes this
23:42video, post a comment, and subscribes to
23:43the channel, it's going to win 1 million
23:45buck token. All right, let's pick the
23:47winner from the previous
23:49video. And the winner is, "Thanks for
23:52your invaluable help and education. Best
23:54wishes. Thank you so much for your nice
23:56comment. Please reach out to me so that
23:57I can send you your bunk tokens.
24:03[Music]

About this transcript

This page contains the full transcript of My Python strategy with 100% win-rate (+ live results) by Algo-trading with Saleh, generated from the public captions YouTube serves with the video. The transcript has 4,415 words across 606 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 translate the transcript into 100+ languages.

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, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.