Algorithm Engineer Reveals The Hidden Forces Behind 'Stop Loss Hunts'

Full transcript with timestamps

0:00 So ju just describing that price signature of a predictable area stops are likely below price goes grabs the stops and then goes in the direction they predict any stop loss hunts. It's a dynamic I've explored a lot with many guests and I never come to a clarity of what's going on here because it gives the illusion that there is a higher power in the market and they are coming to the inferior and and and exploiting them. Yes,

0:24 if we are to assume that that premise, is it even worthwhile to then for them to put the liquidity to move the price to grab the crumbs what it seems of what those stops could be? What is the mechanics of what is going on here? Because I see the signature all the time. That's an question. Here's the answer to that. I wondered about that for years.

0:43 Those advantages are taken by trade placement uh algorithms. So now you ask what does that mean exactly? So now imagine that you um you have um how do I explain this? Let's say that you are a person that's writing an algorithm that is going to do VWAP volume weighted average price and you are given um a buy order for a large number of shares and it's going to last a few hours. What you're going to try to do is is if you have a smart algorithm, you're going to try to find those regions where there's this weakness and you're going to try to use those to do more buying

1:28 because since you can be relatively certain higher than 51% whatever that some of the triggering of the stops is going to produce the price going down, you can use that as a way of getting a better price than you otherwise would. So, it's a way of improving your execution. who who's creating the cause like the cause is somebody an institutional guy that wants to buy million shares of IBM because he wants to invest in IBM

1:53 right he doesn't care about stops and things going up and down and all the rest of it but there is that demand but that demand is not all going to come in at one moment

2:00 so is is it a a matching algorithm to minimize slippage yes it's a and to get the better price and so you know that you could that if something is collapsing during that period of time you're getting a better price than would otherwise. So get in there and take advantage of it as much as you can.

2:18 And and therefore, how do they engineer it? Because the logic makes sense. There's a bunch of stops. I got a guy with a big order. Might as well drive down price. A lot of things happen. Liquidity is found. But who is creating or engineering that?

2:30 I can tell you how we did it. When I was helping Joe with his u um trading algorithms, what we were doing is we were actually using what we call trader logic, Joe's term. And the idea was to do exactly that. Write in a computer program what a trader does as visual identification of you know areas of supply and demand or areas of support and resistance and so on so forth. And basically say we're going to adjust our trading algorithm where you know we're getting orders from customers and we're going to adjust the trading algorithm to take advantage of these particular periods where something interesting might be happening in the market. So another one for example was I mentioned earlier was opening range breakout. So you knew, for example, that statistically speaking, if there was a breakout from the opening range, the stock was likely to go that way for the rest of the day. So you speeded up your

3:19 VWOP on the all day VW. So I I want to speak to you in my language and hopefully you understand, but I'm a purely technical trader and I'm down in the lower time frames. I'm in and out within a couple of hours.

3:29 Y and what I'm looking to exploit is what we're describing here. Market open where I see more liquidity, more orders. And then I'm looking for predictable areas. Could be an Asia low range. Could be a support level. I love take taking out a demand area. When I see that is getting swept and then in the right time, usually open, I see a sense of confirmation. Could be as simple as a break of structure in my direction. Y

3:50 I'll look to long uh stop loss at the low of that sweep and look to a session high or something like that.

3:55 Y and very easy or very often I can get a 1 to three or 1 to four risk reward in the space of an hour and that became my system.

4:02 Y can you help me understand the why of what I do because I've seen it a million times. So, I can believe in it because I've tested it, seen it, but I I rarely understand why it's working and therefore the valid losses that I take that optically appear the same. I don't understand why it was a loss. I just accept it was a loss. Think of it as an iceberg. Um, basically, when there's an institutional order, what you see is some number of shares, but in fact, there's a giant iceberg below them

4:31 that trigger upon the first one. trigger upon the first one and that's why you get the drift in that direction and all that's happening to you when you get losses is that something else happened and either you made a mistake or you something else looked like the iceberg but it wasn't the iceberg.

4:46 What you're taking advantage of is you're saying listen I'm the uh you know what is that little fish that eats the the parasites on top of a shark.

4:55 Uh I know you're for let's say like a leech. Yeah. Right. So that's the idea. So that's what you're doing. Yes.

5:00 You're basically saying, "Okay, I know there's a whale out here. He's going to be trading in this direction the rest of the day.

5:05 I want to be trading in his direction and I want to just basically whenever I can take some Yeah. Exactly. catch a ride." And what you're trying to do is you're trying to identify when that ride might occur.

5:14 And what you found in your trading rules is when that ride might be occurring, presumably more often than 50% or whatever. And you're basically making a nice profit from it. And one thing you could do, by the way, to help with your um P&L is don't put your stops at the obvious places. Am I not safe to assume that the area where the predictable stop losses were that then got taken out that for some time that should be the protected low of the day?

5:43 Yes, that's great. And therefore, is that an obvious place to put it or not? That's not an obvious place to put it. But an obvious place to put it is um you know, here's the market going like this. here's a, you know, line I can draw across the lows support level, fine. I'm going to put it, you know, one tick below the support level. Bad idea. Don't do that.

5:59 Or if it's a round number, please don't put it on a round number. Okay. What's the story though? Because when I when I try and understand it further, I feel like I'm getting into conspiracy world where they are they are hunting the stops of the retail.

6:14 No, they don't have to hunt the stops, but but visually that seems it seems that way, but they don't have to actually hunt the stops because they know they're there. Mhm.

6:21 And so what they're doing, but are there not stop losses everywhere in the market or they are concentrated?

6:25 Yes. Because people psychologically put them at specific levels. And this is I'm I'm really going to explore this because I with every other guest I never get an answer.

6:33 These predictable areas are retail predictable areas and therefore are they not a drop in the ocean in terms of the stops and liquidity that

6:41 Yes. But the Oh, that's a very good question. So let me say two things to you about that. Um uh fractional shares if I forget, let's get back to that. But let's talk about that issue first. So the thing to understand is that the liquidity of the stock market at any given instant in time is not very big.

7:00 So an actual order, an actual retail stop order can move the market even though you wouldn't expect it to because at that moment in time there isn't much liquidity. Liquidity exists over periods of time. At instance in time it doesn't really exist. I forget there's a guy Olsen and associates way back he used to call it a camel going through the eye of a needle.

7:27 Okay. So basically the camel is a big order and you got to push it through the eye of the needle at at the eye of the needle itself has very little liquidity.

7:35 Mhm. That that's why those orders matter. And the second thing you want to do, by the way, is when you're putting your orders in to go in the direction of the flow of the of the whale, please don't use round numbers ever.

7:49 Mhm. Use something really stupid. Make yourself look like an idiot. So, put an order for 96 shares or, you know, 221,

7:59 some really crazy number. Use prime numbers. That's even better. um that don't end in five because prime numbers and also numbers that don't end in five and zero. when I this signature that I'm describing where we get predictable area sweeps it

8:12 yes bounces up gives me a confirmation I get in y a lot of the times which is happening more now than it ever used to y I wonder why

8:21 that protected low that is ordinary a protected low for me gets swept once again so it comes into a where the stop losses should be taps once taps once again and then goes

8:31 yep this didn't used to happen I didn't see it 5 years ago or 10 years ago I'm seeing it more and more now what is going on here. Same thing, targeting the expectation is that that area may have some stops. And so what you what you're doing then is as an algorithm, you're basically slowing down your buying. And if you slow down your buying, of course, the stock's going to fall.

8:51 Mhm. Because it's your buying that's been propping it up. And you wait to see if you can get it at a cheaper price by slowing down how you buy.

8:59 Can can this be victim to spoofing? Absolutely. No question. And nothing to be done about it. It the game is the game.

9:07 The game is the game.

Questions about this episode

How long is "Algorithm Engineer Reveals The Hidden Forces Behind 'Stop Loss Hunts'"?

The episode runs 9 min; the transcript is about 1,936 words in 37 timestamped sections.

More from Titans Of Tomorrow

• All analyzed episodes