218: Trading Triumphs - Jerry Parker #3: The Dangers of Optimization

and the unexpected impacts it had on his trading results. We're gonna talk about over optimization, how we can avoid it, and some uh solutions for traders, plus a whole bunch of other stuff as well. Jerry doesn't hold back in this one, so let's get to it.

Uh mm it may make it'll make money, but it may not be a lot of money. But it really um you have to sit with a lot of drawdowns and a lot of whipsaws, but the one thing it does give you is it's very reliable. And so We were just using these simple approaches and having so much success. But it kind of like went to our head a bit in the sense that well, let's do better.

And you know, making money every year, that's pretty darn good. And our clients loved it. Clients were very greedy and they were throwing money at us because we were so consistent. But we couldn't help ourselves and we tried to do research and And it was basically trying to improve things. We were listening to what other traders were doing, profit objectives, take profit, reduce your positions based upon volatility.

Abandon that. That was another good thing too, that it doesn't always happen with traders. I've seen traders get into the same situation and be so committed to backtesting and research, they're almost obsessed with it. That if they turn their back on their latest research or the whole process of doing research.

then it's they really can't handle it. And we were basically came to the opposite conclusion. We were saying we just w w went too far. We can add more markets and we can become more diversified. Maybe we should be longer term. There's other things we can do in our research process that doesn't violate The Ten Commandments of Trend Following, all the great rules and philosophy we were t I was taught. And so

I think that was a good move on on my part to sort of realize this incredible mistake that I had made. Because I literally was looking at just months worth of data that Um I probably we had lost money four months in a row and our relative performance was pretty bad. And we had never seen something like that in the back test. So that's a very good sign. You you look at this back test and right out of the gate.

You know, the performance is so bad and it doesn't look anything like this wonderful back test. Go back to the drawing board and be willing to s to swallow your pride because it's all about making money. And doing what it takes to make money. And it's not about scratching your itch.

to make sure to make you think that how great you are as a back tester or how smart you are. I like to think that I'm smart, but I would prefer to trade a methodology that's a little less involved and less complex if it's going to increase my ability to make profits.

How how short-term can we trade? We don't want to be too short-term, but what's a good short-term parameter? So we would test that out. Uh once again just one entry, one exit and a stop loss. And Then we would say, okay, how l how long term can we trade? And then we would get a pretty good indication of how that worked. And then we would literally just say, okay, just trade a a couple systems in between without really even caring about optimizing the parameters so much.

Um another way that we get away from optimization is we don't pay very much attention to the equity equity curve and the equity stats, the drawdowns and the The volatility and trying to smooth out that equity curve. Whenever you try to smooth things out and increase your sharp. That's where it all begins. That's where the uh over optimization begins. All we paid attention to really was the trade stats, the average win, the average loss, the average trade, the win percentage.

There's a hell of a lot of stuff that happens in between the entry and the exit, right? We just ignored it. We were like, we don't care what happens in between. We just want to find systems with healthy um Trade stats. and that make the most amount of money um per uh you know per small loss. So the the win loss ratio. Here's what the here's what the average win is, here's what the average loss is. Let's try to maximize that. And close our eyes and live with drawdowns, live with volatility.

But a trade leverage wise in a way that makes that possible, of course. But I think uh and I've often said over the years, I try to take as little from the back test as possible. Here's a good place to buy, here's a pretty good place to sell, we're done. And um One of the things that when you get into this trend following uh the way we do it. When you get it, you'll see that there is this

um group of parameters, the shortest to the longest, where um this all the systems make about the same amount of money. And to me that's very company. Um So yeah. I'm trying to give you lots of examples in verbiage, but really I try not to use research very much or backtesting very much. And and these are the reasons why. It's totally unnecessary. We want our systems to

It'll help you over time get to a place where you can do more. You'll make money as long as you have a good medium to long-term strategy that doesn't try to Get in and get out too quickly. I used to say we take small losses. Now I think it's an optimal loss. I think all of these. uh entries and exits and parameters are kind of opt optimal. Um there's nothing magical, it's nothing special. Really what makes them special, what makes them work. Is the is is always doing them.

is um always following your system, whether it's one system and maybe later it'll be two or three or four. But don't do do whatever you can to put yourself in a situation where you can do all of the trades in the right way, in the systems way. Um If you have to make this cutback, the great thing about the cutback too, if you have to trade small for a while, is It allows you to keep doing the trades. Whereas if you sort of like say well I I'm really on the defensive here.

I need to trade smaller. I'll I'm gonna have to choose some trades not to do. That's what you don't want to do. You wanna do all the trades in the way you're supposed to do them. And this cutback rule says, oh, all we did was just reduce our leverage, reduce our our current positions and our future positions for a while, but we can go right back to doing all the trades we're supposed to do. This is magic. This is uh a superpower.

I've been consistent. Our our systems are very similar and they're very correlated. And from a norm from a traditional way of measuring correlation, people might ask, why do you even bother? They they're all long. You know, you're long orange juice or cocoa and you've been long for a year or two. They've all been long for a year or two.

And I'm like, yeah, that's wonderful, because they're all making lots of money. And so the correlation measurement is kind of a trick because it doesn't really matter so much that in these trends um where it really matters, your exits are really gonna have a big impact. Um correlation is the wrong measurement.

the when you when you finally get out of the sugar and the cocoa and the yen and some of these and Eli lily, these exits that you have on your systems will more than likely produce vastly different PLs. Some will get out near the high, some will give back more profit.

Some might get out too quickly and then the market will will go back to the highs and keep going and the longer term systems will do better. So it's really for a trend follower looking at the value of diversifying the exits in terms of the profit in the trade. You sort of um no you you don't know where the perfect place to get out until after the trade is done. And then when you look at all the data, forty, fifty years worth of back testing, all the markets.

The conclusion that the computer shows you is all these systems make about the same amount of money. Of course they do, because you can't predict.

They kind of these breakout rules, this is this is kind of a function of using breakouts. But um once again Um, that brings up another idea of how important it is to be diversified so and trade as many markets as possible. If you're getting them all at at the same place in the crude.

the entries in the heating oil will be a bit different and more spread out. Uh it's very important too to um Not try to manipulate the systems and the entries and the exits to try to achieve something else like this diversification or spreading your risk out. you it just leads you to trying to um not really follow the system as much as you should. Um so I think I have done that as well. Um I want to space it out. So I'll force these systems to space out in some sort of way that.

makes me feel better because I don't like it when they all get in on the same day. You're screwing up with the power of always doing the same thing every time, which is primo. That's the most important thing. Yeah.

All right, well um so how do we summarise all that, Jerry? You you sh we were just talking about optimization and y shared so many different uh uh insights here. So um how do we summarise Uh how do we summarize that?

And then we'll have an exit um in the same you know, to get out with a breakout. But um yeah, we don't want to add in more and more filters, more and more uh conditions. uh loose pants, let the markets go, have multiple systems. That's rather than overloading your one system with all these different rules, have multiple systems with almost with very few rules. That's how you s stay on the right side of the robustness gods. They will punish you.

Um w when you have the opportunities, your job is to do the right thing with the opportunities. It's not to uh you cannot produce these big trends. The one thing that you need to be successful is totally out of your control. The only control you have is did I do what I should have done? With my possibly mediocre system. But a mediocre, I'll take a mediocre trader, a mediocre systematic approach.

That's not overoptimized and a person who's dedicated and willing to do all the trades and hang on to those trends and not be eager to get out too quickly. uh over a more sophisticated approach that can break uh much more easily. So I am so thankful for these markets that we have and how they have bought all the houses that I own and all my boats and all my standard of living just because

uh I hung on to those long term trends. Whenever I talk to traders, that's the one thing that they all tell me is I got out too quickly. I think um you wanna be that person who Try to stay as long as you can. And you only have to stay as long as your system tells you to. You can uh it's not like you have to do something that's not correct. We only do it because the computer tells us to do it.

Well what a way to end our discussion, Jerry. Thank you very much for that. Actually before we go, I have to ask you, what's that picture behind you there? Uh that's the

Um we do a trend following spaces every Friday, seven AM New York time. And we have some aussies on there. So I know it's late. Yes. Nine or ten o'clock at night, but I did a I did one of those spaces from Uh from Australia actually, late at night. So it can be done. And we have some good friends on there. So that's a good place to learn about trading and and uh to have some camaraderie.

about trading. November was a tough month, so it's good to have people to um we're going through the same thing sometimes.

the call, it was really nice to listen back to. So I'll have links to all that on underneath the video here for anyone who didn't catch it just then and go and uh check that out. So uh and also if you enjoyed my chat with Jerry and everything that Jerry shared with us today, please give us a thumbs up on the video and don't forget to hit subscribe and the bell button to be notified of any new content.

All right, Jerry, well thank you again for today. It's been an honor speaking with you and it did go by so fast. It was really enjoyable. Any closing comments?