Does your last trade have anything to do with your next trade?
In this episode of 2Traders Podcast, Walter talks about the “gambling fallacy” mindset, back-testing, and statistical analysis; and how these can affect your decisions in determining trading results and the amount of risk you are willing to take.
The “meat” of this episode centers on whether any individual trade taken is closely related to the previous trade or not.
Darren believes that while past data on charts and trends might be a great source for future expectations, it should not affect the next trade that you are in.
Sometimes traders run into a losing streak and think “This can’t go on forever, right?” But what if it does?
Finally, the conversation wraps up with a discussion about roulette and blackjack in Las Vegas… and how the casino’s point of view may help you (seriously) in shaping your trading strategies.
Download (Duration: 20:27 / 23.4 MB)
In this episode:
01:43 – surviving a losing streak
02:03 – gambler’s fallacy
03:35 – complete randomness
05:43 – series analysis
08:18 – magical trend system
10:22 – what creates my losers
12:14 – being comfortable with your system
14:11 – figures in trading
20:01 – the observer effect
Tweetables:
How does the “Gambler’s Fallacy” affect your trading? [Click To Tweet].
Don’t use past trade data to anticipate the future. [Click To Tweet].
You can be right and still go broke. [Click To Tweet].
Download The Full Episode 34 Transcript Here
Darren: For me, the movement of a price chart pattern is just complete randomness.
Announcer: Two traders, Darren and Walter, pull back the curtain on profitable trading systems, consistent money management, and profitable psychological triggers. Welcome to the Two Traders Podcast.
Walter: Welcome back to the Two Traders Podcast. Walter here, I’ve got Darren on the line. How are you today Darren?
Darren: Hello Walter. I’m very good. Thanks, mate.
Walter: Well, I want to talk about the gambler’s fallacy. I’ll just introduce it for those of you who haven’t heard of this idea. So my friend and I in university, we had a lot of statistics courses. Mostly because they told us that if we were to take a lot of statistics courses we’ll be likely to get into graduate school.
They said, “You really need to take statistics courses.” So, we take a lot of these courses. One of my professors was a fascinating guy. We had this idea, and the idea was that… Where I went to school, Darren, wasn’t too far from Las Vegas; it was maybe, like a four-hour drive from Las Vegas or five-hour drive from Las Vegas. So we had this idea that we’ll go to Las Vegas and play blackjack and we’re going to make all these money. The idea my friend had was that after every bet if he lost he would, of course, double the bet. If he lost ten dollars, the next game he’ll play for twenty. If he lost that, the next game he’ll play for forty and so forth. He keeps going up.
He had this deck of cards that he carried around and he kept playing the game. He wanted to see how many losers he can have in a row because he wanted to be able to survive a losing streak.
So, we presented this idea to our statistics professor. He said, “Look, why would you just do this if you went to Las Vegas because you’re sure to make money? Eventually you’ll win and you’ll get that money back plus a little bit more.” He looked at us and said, “Are you serious? That’s the gambler’s fallacy.”
Which is basically this idea that if something happens more frequently than normal, it’ll happen less frequently in the future. Same thing with roulette. If you’re looking at the roulette wheel and you see that red has come up twelve times in a row, you would go down and put money on black because it’s certain to come up again. It’s due to come up again.
So, that’s the idea. These sorts of things are related and that they’re not independent events. My question is: how does that relate to trading? What are your thoughts about this idea of independent trading results versus related trading results?
Darren: Yeah. Well, I’m very fifty-fifty down the line. Its all random. Whatever happened before has no relevance on what’s happening next. The same with my ideas on technical analysis, really. You’re dumb if you do. You’re dumb if you don’t, really.
I mean, is this the same certain notion where all these trends been going on for way too long so it’s got to reverse now? Because, it has this greater chance that it’s going to reverse. Where it’s actually, it’s just as likely to continue falling as it is to reverse.
There’s this human behavior element coming to it then and affect that. Because for me, the movement of a price chart pattern is just complete randomness. But obviously, what’s making that occur is not randomness. I’ve kind of a bit modeled with it really I mean.. Is that what you mean or I’m way off the mark here?
Walter: Yes. It’s just the idea like…A trader comes up to you and says “Darren, I’ve had seven losers in a row. My back-testing has shown that it’s highly unlikely to have seven losers in a row. Then it’s a 0.5% chance of that given the parameters of my system. What do you propose that I’d do for trade number 8? Because I’m thinking of increasing my risk for that trade,” right? So what would you say?
Darren: Well, I would tell him that his trade is fifty-fifty. No matter what has happened before, it has no effect on the next trade you’re in. You’ll be very unwise to load up on this next trade with some notion that you’re losing streak has got to end cause its just not true.
Walter: Absolutely, I think it’s the same thing in terms of traders who’d say… I remember the first couple of traders that I was showing them what I was doing. It was sort of like this mini course in my living room. Then he said, “Tell me some money management or risk management strategies.” One of the ones that came up with was, “We would risk more money on those trades that we are sure to workout.” I thought that was bizarre. Are you serious?
You’re going to risk more money when you know it’s a winner. Why even risk any money at all on the ones that you don’t think are likely to be winners? What’s the point?
It’s the same idea here. I will say this, in fairness to those people who are thinking right now, listening to this thing: ‘Well, you guys are getting it all wrong. There are some trading systems where each individual trade that you take is not independent and it is related to the previous trade’. That’s true.
There’s a whole area of statistics called “time series analysis”, where they’re looking at different events as they occur in a series and they’re related. The rules of statistics totally changed because in all statistics, you assumed that every single event is independent. If not, then you’ve got to move it to this other box they call, basically, “related events time series”.
Here’s an example, Darren. Let’s say, I’m a trend follower and I’ve decided to go short the EUR. I’m going to sell the EUR. I’m going to buy USD because it’s in a downtrend. Well, if this is the case, then every time I get the signal that EUR pulls back, comes up a little bit. I get a signal to sell it again, I would sell right? Fair enough.
Well, what’s going to happen at the end of that trend? When that trend is over, and the market decides to start going up again instead of down, I’m going to start to lose money. What that means is, because the market is starting to change, maybe my last trade was a loser because I thought the market was going to make really low and it didn’t. Then I took another trade because I get another signal and that was also a loser. And I took another trade and that’s also a loser. So, I’ve taken three losing trades in a row, and they really are in some way related because the market is no longer in a downtrend now.
What I’m saying is, if you’re a trend trader and you’re taking trades in your biases in one direction, then you would expect some correlation to be involved when you’re looking at trades around that recent trade that you’ve taken. Does that makes sense?
Darren: Yeah. I suppose. Are you saying that if you have a streak of losses and you’re a trend trader, then that’s a pretty good indication that the trend is failing basically. A weakling. Is that what you’re saying?
Walter: Yeah. That’s the reverse. That’s the other side of the coin. I would expect that at the end of every trend, I would have a lot of losers as a trend follower. Also, I would expect that my losers would tend to clamp together.
If I’ve got this bias and I’m only taking buy trades or sell trades based on what I’ve determined as the trend, I would expect that to happen. In general, I’ll have a lot of losers clamped together. I have some sprinkled in there of course, during the midst of a strong trend. But mostly, they’ll clamp together at the end.
By definition, my trades aren’t really independent because losers are likely to follow losers. Whereas, in general, that’s not to be the case. Now, I just made this magical trend system up here but I’m just talking it out loud to show why some traders who are listening to this were thinking, “Yeah, you’re right.”
Like, Darren, if you’re trading the GBP today and tomorrow you’re trading the EUR/CAD. Then the next day, you’re trading CHF/JPY. Well, why would they be related? They’re independent trades. They’re independent markets. Even if you are using the same strategy, why would they be related? If I’m trading the GBP straight for two months then I would expect my losers to clamp around those periods when the trend is starting to fizzle out.
Darren: Yeah. I suppose they’d give you some information certainly but there’s still no reason to trade for. There could be a massive winner and the trend resumes. Certainly, it is great for expectation. Certainly important for expecting what you’re particular strategy is going to give you in a way of streaks of losses.
Last week, I had my first 8 trades of the week were losers. In a way, I continued to trade it because I knew that the possibilities of it continuing was slim. But not because I’ve had those 8 losers. More because it’s unusual for me to do that. At the same time if its gone on a week, I could’ve get it. But I wouldn’t think that it has anything that I could’ve avoided by looking at my other results.
Walter: You don’t use past trade data to anticipate the future. In other words, you don’t say “I had 8 losers…” I guess, in a way you’re looking at it going, “Well, this can’t go on forever”. Right?
Darren: Yeah, because it’s unusual for price to consolidate all week. Basically, that’s what creates my losers. If its gone up a little bit, down a little bit in a day, pretty much where we started. That pretty much nullifies any edge my strategy has got. It’s quite hard to find in a week’s where that happens all week.
Essentially, we’re talking about a very narrow range week in a bar. Having a weekly chart, you do get one or two of them but it’s much more likely that there’s going to be some movement at some point through the week in the currencies that I trade anyway. I’m going to have that expectations that I got a bull market rather than looking at a specific data.
I’ve roughly know, I’m going to get one or two days a week. Most likely two days a week, where my strategy is horrible, but I’m certainly not going back for twelve years where I count the exact amount because I think that’s just a complete waste of time. It’s just trying to be too complex, really, and I think like I always say: “It’s trying to remove too much uncertainty with data.” I know that is comforting but I think in the long run it’s damaging and makes trading harder.
Walter: Right. Let me ask you this: what about to those who say, “Yup, fair enough. I understand that I shouldn’t use this data in my future projections or management or whatever on my trades but I was able to get some experience trading the systems”. Do you know what I mean? I saw different market conditions and I’ve executed over and over again in forex tester so at least I’m comfortable with the system.” What do you say to that?
Darren: Yeah. I’m kind of cautious to anyone’s road to learning to be a trader and say it doesn’t work. I’m really trying to avoid that because I don’t believe you can’t really say that about anyway of trading. I think it’s important. But again, I think it’s given too much importance and not enough importance is given to the fact that, that data is really just a bull part figure and different market conditions can basically throw all of those figures where it [12:48 inaudible]
Even, like minor differences, where two years look particularly like similar movement. Just slight variations in the price action can change that data. I think if you guarantee it too deeply, you start to alter your strategy and you nullify the edge that it’s really got.
Quite often, people’s edge in trading is just something simple like the use of a bit wide stop initially and then move up or it might be that they let trades run. They’re either taking a risk off really quickly getting it start a lot so taking very small losses and when they do, it has nice momentum in that. They got a big win, that’s when profit comes, vice versa.
Like me, I take quite a lot of a fairly large hits but when price finally moves, I tend to have quite a few positions in and I let them grow. On average, my winners are bigger than my losers. I think it’s subtle things like that, that make profit and not like the particular going into the figures of win rate and lose rate in the figures. I just don’t like figures in trading.
Walter: Right. So you’re definitely not a fundamental analysis trader like looking at figures and reports and things like that?
Darren: No.
Walter: Yeah, I agree. I don’t think you can trust what the government numbers are anyway. I think that, some of the most savvy guys that are fundamental traders, they don’t look at official government numbers. They’ll look at other things that tell a story. Yeah, fair enough.
Darren: It’s relative as well, isn’t it? I was on a forum the other day when, I think it was the Bank of England, where the data came out and I’ve kind of know what the data was. Essentially, they were expecting 2.2 or something and it was 1.4 or more of that and instantly the messages started coming down. “The data was bad and price goes up. What’s going on?” Well, who says it’s bad? It was worse than expected but it might still be a good figure, so it’s relative really.
It’s relative thinking. It’s good or bad. Everyone is going to have opinion on it and it’s just one figure. What about all the other figures that didn’t come out today and all the figures for the currency that you’re trading it against?
That was just really a good example for me. How numbers, particularly, can really fool people and just instantly they’d rely on their bias. So numbers out in two minutes, they think that it should make a certain reaction in the market and it did the complete opposite.
Walter: It’s like the gambler’s fallacy. Just because you’ve had so many winners or losers, it doesn’t mean that it has much to do with what’s going to happen next. I will say that it’s tricky though as traders.
For example: if you know, Darren, that your win rates is going to be around 50%, it’s tricky to know that after eight losers in a row you are still trading that 50% win rate. Because, it feels like the market has changed. Feels like something is wrong. Feels like it doesn’t work anymore but you take a longer term point of view and just ride it out.
I have to know. So how did it end? How did the week end up for you? Was it down a little bit still? Or was it a profitable weekend?
Darren: No. I shot on Wednesday evening 50% of my target for the week.
Walter: Right. That’s excellent.
Darren: Monday and Tuesday it didn’t move. Then it broke out and went long all day Wednesday. By the end of Wednesday, my two winning trades covered my eight losses and gave me my target for the week. That’s what I expected to happen. Sometimes it breaks out and it just covers my losses and you just have to accept that.
But it’s difficult. If you think about the figures all the time, I think it becomes a bit of a burden for you. Because you’re comparing your “what is happening right now” with “what you are expecting to happen”. What you’re expecting to happen is based on what happened in the past and usually people are using a lot of data. So that’ll take five years of data and then you know they’ve taken eight trades. They’re expecting those eight trades to somehow replicate the average.
For me, it doesn’t make sense. It’s illogical, but I’ve sensed people go out of way and a lot of people make it work. One last thing on the gambler’s fallacy thing, I read a lot of really successful traders in the past say, “You’ve got to go all in when you know it’s good and then be cautious when it’s not.” They’re falling into that trap as well but somehow they make it work for themselves. Like the guy who shorted the Bank of England or something and make millions.
Walter: Yeah, Soros. George Soros.
Darren: Yeah. Is he using that book? Kind of making it work?
Walter: Yeah. From everything that I’ve read, Soros is basically using it and has taught to use that money management strategy that I was laughing about when I first taught trading to a couple of traders. He basically says, “It takes courage to be a pig and you’ve got to go all in when you know you’re right.”
Maybe, he was shorting the GBP. I think, he was shorting the GBP against the Deutsche market at the time. And I think maybe he was shorting it, shorting it and shorting it and then finally when he got something really good, he just went all in or something.
Maybe he was Flick in the Big Short when those guys were right about shorting the housing market and the mortgage market. They were right but they still had to ride out the storm. They were losing so much money and they had to pay into these trades just to keep them open. Even though they knew they were right, that’s one of the things about being over leveraged and risking too much. You can be right and still go broke.
Well, I’ve really enjoyed this one, Darren. Thanks for your thoughts on the gambler’s fallacy. What are we going to talk about next time? Any ideas? I’ve been thinking about a few things.
Darren: I’m short on ideas. I find it very difficult to fit into a trade. I’ve been constantly searching for someone who’s saying the same thing as me and I can never find it. That should worry me but I suppose, really, that might be a good thing. Really.
Walter: Yeah, I know. I think that’s a great thing. Let’s talk about this idea that I think is interesting as it relates to trading as much as this idea of an observer effect. So maybe next time, we can sit down and talk about the observer effect and how that might affect your trading. I think that would be an interesting one.
Darren: Cool.
Walter: Alright. See you next time Darren.
Darren: Sure. Cheers, mate.
Walter: Bye.
Mark
Hi guys,
I think you should revisit this topic.
I understand the concept of the gambler’s fallacy but it doesn’t make sense.
For the sake of my argument, tossing a coin gives a 50:50 chance of coming up heads or tails.
If you hit a string of tails then the more tails you get the closer you are to getting heads because you aren’t going to gets tails indefinitely. Maybe in theory you could but in practice you won’t.
If the gambler’s fallacy was true then you wouldn’t have a 50:50 chance of heads or tails when tossing a coin, you would have some other probability.
In this case, either the gambler’s fallacy is true or a 50:50 chance of heads or tails is true – they can’t both be true.
If traders believed the gambler’s fallacy then they would stop trading every time they hit a losing streak.
Instead, successful traders embrace the gambler’s fallacy every time they continue to trade through their drawdowns because they believe that with every losing trade they are coming closer to their next batch of winning trades and that their trading strategy will bring them back to their win/loss equilibrium.