CLICK HERE TO MEET ADAM AT THE LONDON CONFERENCE
In this episode of Truth About FX, a special guest joins Walter to talk about automating your trading systems and what it’s like to be trading for a hedge fund. Adam Hartley is a long-time forex trader and programmer with a Doctorate in Physics. His background gives an interesting twist on how he treats his charts and developing his strategies through algorithms.
And, you’ll here about a good tip on how to spot a true ECN. More on this episode…
Download (Duration: 40:39 / 46.5 MB)
In This Episode:
01:21 – hanging over you
03:23 – globalization
05:57 – purely mechanical
07:57 – bad patch
10:24 – dark art
24:49 – big, impulsive moves
28:12 – accelerated
32:54 – purely mechanical
35:35 – trend biases
38:46 – how to spot a true ECN?
Tweetables:
Make a living in doing what you love [Click To Tweet].
Those occasional winners can make the difference [Click To Tweet].
Balance the risk/reward against the hit rate [Click To Tweet].
Announcer: Sometimes, forex trading can be a wild and wooly place to be but, forex trading doesn’t have to be the bad lands for your partner. Instead, hunker down and bend your ear. You will be burning the breeze in no time here at the Truth About FX Podcast.
Walter: Welcome! It’s Walter here. Today, we have Adam Hartley. Adam has been programming for over 20 years. He has a Doctorate in Physics and he’s got quite a background in programming. So, it’s going to be an interesting conversation because Adam is going to shed a little bit a light on the Algo Boogies, perhaps, and what it’s like to work for hedge fund. So, welcome Adam.
Adam: Hi there, Walter. How are you?
Walter: Great! So, Adam, this is really interesting because there aren’t a lot of people like you out there who are willing to talk. I am just curious about your background. How is it that you went from Physics into, essentially, programming algorithms? How did you make that jump? I guess, I suppose what’s your story like? How did you get to where you are today?
Adam: Sure. Well, as you said I started off in academia. I was studying Physics and during my Doctorate in Physics and there’s always this thing hanging over you: where the next bit of money will be coming from as an academic and I was starting to worry what will I do if the funding ran out because I was enjoying the Physics.
So, I started out. I’ve always been quite interested in the market. I started to think, “Well, maybe I could sort of figure out the market. How hard can it be?” So, I started to get interested in it, read some books and started to code up a few little simple indicators, really basic stuff back then.
I just really got into it. I got more and more addictive to it. I wanted to do work, hard work, didn’t find it easy as I thought. I tried to do a little bit of spread backing back then with predictable results losing money but I thought there must have been somewhere who is doing this.
As expected, it didn’t go well and then the money did ran out and so I took a programming job with some guy in one of the neighbouring department. I set-up a little spin-off company so I did that and I was bored rigidly. I started developing various contacts in the finance world and I eventually decided that I really like to work for myself.
I’ve got my programming skills, I’ve got my interest in finance. I’ve discovered program complaint station which is very good for programming stuff and testing, backtesting. Also, I’ve been developing a little charting package. I’ve got a little handheld pager, got a futures pager which will sort of update the market every sort of 30 seconds to a minute.
I wrote an inspace to it so I wrote this simple charting package. Eventually, it’s around 1994, I took a plunge and started working for myself. I’ve been so busy. I’ve been doing that ever since. My sort of clientele vary from a sort of retail traders all the way up to the biggest institutions.
I’d do whatever people want me to do, whether it’s just programming up an indicator or selling my charting stuff. Sometimes, I was doing serious backtesting for large funds and things and I started doing that and I just progressed.
I’ve made a comfortable living in doing that and I care on doing that for quite a number of years. Then, gradually, things started to get a bit harder to make a living. Partly I think, because of globalization. There is lot of really cheap, good programmers in some of Eastern Europe and Asia who came online who do it for far less than I would so the work started to drop a bit but then I met a guy here in Oxford.
We just started to develop trading strategies for a fund and he was looking for a programmer so I came on board and joined him. It turned out he was working for the Tudor Group, Paul Tudor Jones’s fund.
We worked together developing strategies for Tudor and that went on for 10 years. Actually, it was a very successful collaboration. So, we ended up designing from a really quite large scale trend following, one term trend following strategy, which Tudor traded. They traded over 800 million with our stuff and we get sort of a small performance and management fee.
That worked very well for a long time and then, unfortunately, the trend following hit a rough patch towards the end of the last decade and the relationship with Tudor came to an end so I went back to doing the research and programming, things like that. Also, I’ve always wanted to trade.
The thing that actually got me into this whole business is, as I’ve said in the beginning, was I wanted to learn to trade so I can finance myself. So, I’ve always been on the side trying to work out of trade, scout the Naked Forex Now Forum, and got really involved in that and it was great to learn all those sorts of different, new strategies from that, meet up all sorts of fellow trader,s and things and stuff. That’s been one of the things I’ve been pursuing alongside with programming is trying to get in a stage where I’m, actually, I’m a professional. So, that’s basically where I am now.
Walter: Okay. One of the interesting things that I’ve heard you mention is — I think it was on the Forum, the Naked Forex Now but I’m not certain. I think that’s where I’ve read you were saying that you’ve found that it’s easier for you to program an automated system that’s a trend-following system than some other sort of systems.
So, I suppose if that’s true then, you’ll be comparing that to like a mean reversion or a breakout or something like that. Can you explain that because I find that to be an interesting comment? If you didn’t indeed say that.
Adam: No, I didn’t say that and I’ll explain that. I found when working on some of the Tudor years as it were, we developed a lot of strategies. I mean, there are some constraints that there has to have big enough to manage millions of dollars so we cannot do small finicky stuff.
It was very hard to find something purely mechanical that’s purely rule-based which would actually make money consistently. We found that and, very often, we developed some idea and you tweak the parameters and interrupt and to let it find out where it works best.
The system will always work in the end in the trend-following one. The reason for this is one of the most enduring engines in the market is a long term trend. I mean, if it trend over the past six months then, relatively, it’s easy to make system which will exploit the edge and will make money. It seems like those in other edges come and go.
One of the most consistent edges which will endure the tend to be trend following and… That’s what I know, that there are all other measures out there but they tend to come and go. It’s so very hard to get something that works well consistently over a long period of time.
When I’m talking about a long period of time, that’s 10 tough years, sort of thing. So, basically, it’s hard getting developing purely mechanical stuff. One of the few things that will work is trend following. It has to be done in a long term timeframe which is probably far longer than most people who most retail traders will be more interested in.
Walter: So, when you say on a long-term timeframe, you mean the system is programmed to work say on a monthly chart or something like that. Is that what you mean or you mean is there something else you are are referring to like a 100 day lookback or something like that? I don’t know.
Adam: Something like a 100 day lookback. It’s that sort of order of bank few days, actually. It doesn’t have to be on a monthly timeframe. Most of us process on a daily time frame. It would be for a big lookback and holding trades full to be load for several months sometimes. That’s what I mean by that.
Walter: Yeah, interesting. I recall reading — I think it was Richard Dennis — and he had had as Richard Denis says, as we all know is in the “Market Wizard” book is I think the Turtle Experiment and all that. I think he had some sort of bad patch and they interviewed him.
He was interviewed again after the “Market Wizards”. It might have been in the second “Market Wizard” or something like that. The point was that he said, “Yeah, we’ve found that we had to change time frames.” Like, we have to go much farther out or go much lower timeframe because, essentially, everyone is basically trading 21 day breakouts or something like that that copy what the turtles are doing or something like that.
I don’t know exactly but I recall he was saying since we have to drop it down and look at the lower timeframes or go zoom way out. Is that something that you found also? For example, can you trade — I don’t know — 200 candle lookbacks on a 5-minute chart or something like that? Or, is that not something that you’ve looked into? I’m sure, I suppose, you had plenty of time to play with a lot of data when you’re back in the Tudor days. Right?
Adam: Yes. We look on sort of things. We didn’t generally look at 5-minute charts just because of the issue they are trying to put on the 50-million dollar worth on a 5-minute chart. I admit it was all because of the market but still execution was a big deal.
That’s one of the things that people don’t realize is that when you’re managing 800 million, actually trying to put the position on when you think the trend has changed, it becomes an issue.
Tudor has this rule that that they wouldn’t trade the market if the volume they’re trying to do is more than 15% of the total volume of the day. But, it was approaching that so that’s how big some of the sizes were.
To answer your question about the moving sort of sweet spotters that were in the optimization, that is unfortunately does seem to be the nature of financial market. I mean, perfect world, we better model these things and there’s one ideal look back and you trade it, you optimize it in the past, and it works well, and it stays there evermore going forward.
We won’t be millionaires of that one case but, unfortunately, this sort of optimal parameter is linked that you may be trading for trend-following tends to move around. Just like Richard Dennis found and there’s an optimizing fixing parameters for a system is bit of a dark art, really.
One of the conclusions that we basically came to is that the first is the few parameters you have, the better. Because, this flagbearer of is such a huge thing. It’s far more than you think so with that idea, you have one parameter and even then we try and do to diversify cross by that one parameter.
So, if the trend length will optimally say 8 months or something like that, we might trade 3 months, 6 months, 9 months and 12 months and just try to remove using the dependence on that so, you’re trading huge combinations of different trend links. We’re trading 60 to 80 markets so it’s a huge matrix there of things being traded.
What happens is what emerges from this huge conglomeration of orders is that there is a small but persistent age which you find at the end of the month there is a few percent that you grind it out like that. But, yes, the sweet spotters as told are certainly moved around overtime.
Walter: Right. Do you find that changing the exit parameters or the exit rules is something that has been fruitful regardless of the entry? Like for example, you have those five different entries — or whatever hundred-day high or whatever 500-day high, whatever — something like that and then does the exit make much difference with in your results, in general?
Adam: Well, the kind of stuff that we were typically doing especially for long-term trend following. So, another turtle stuff you actually do, you have one link that’s going in like 20-day and something like a 10-day to exit so you do actually have a clear view that you were flat. The kind of stuff we will be doing will be more like you’re going on long like a 20-day breakout and you reverse short of the 20-day breakout so there wasn’t really so much of an exit parameter as a reversal parameter. Which is like, when you’re not long, you’re short for the thing
That tended to recently work for that particular type of strategy. I know that there’s a school of thought that the entrance don’t even matter and the exit is not important. I actually think — I do subscribe to that, certainly — one of our trend-following things the entry was virtually random but it’s how you manage the trade which certainly made the difference.
If you’re trying to capture a long term trend, you make sure that you got out very quickly if the trend is against you but you hold it for up to 6 months if it was going you favor, sort of thing. There is that aspect to it.
Walter: Yeah. I need to record that and send that to some trading students because they argue that point. But, I think that’s interesting because what you’re saying, essentially, is it makes a lot of sense to cut your losers quickly and really try and hold your winners as long as you can. I mean, really, that’s essentially what you’re saying, isn’t it?
Adam: Yes. I mean, it is what I’m saying for that particular type of strategy and I think so many people that is a very successful way of trading but it’s also many people find it very difficult because the psychological aspect are against you and tourists, we tend like losers.
Instead of cutting out losers short, we tend to do the opposite and then we also don’t like. We like taking our profits quickly because they are on the table. Psychologically, we’re biased against both sides of that equation but it is a good way of trading, in general.
There are many successful people… If you come across a guy called Chris Capre, he has a system called “2ndSkies Forex”. Basically, he stars all about that. It’s about risking 1 hour of his routine, taking 4 or 5 hours, and cutting losers short and let his profits run. It’s a great way of trading if you can get it to work for you.
Walter: Yeah. It’s interesting because I think you’re right, you’re absolutely right. Psychologically, we feel more comfortable adding to a losing position because we want to be right, not because logically the math works out. If you get rid of those problem children and just hold on to the ones that are behaving and, eventually, you will have a great family, right?
But what happens is, oftentimes, we just want to add. If it was a good buy at that price and it’s falling a 150 pips, that’s even better by now. We just want to be right. It’s a difficult thing to break through. Is that something like, it sounds like that’s, essentially, you’re hinting at in terms of the psychology of the trader.
Adam: In terms of psychology, yes it is. It’s exactly right. It’s something very alluring about doubling down on a loser because you then have to go back hard as much now to at least scratch the trade if it keeps going and it’s not doing very nicely. It’s a road to hell, it’s paid with good intention to pay. It’s a bad way of trading, in general. The thing about trading that’s so fascinating is our psychology, in general, is there against doing what is right for trading. You have to do the opposite our instinct, in general tells, us. It’s challenging.
Walter: Yeah, it’s exactly right. That’s why, as a new trader, your gut instinct just going to always tell you the wrong thing to do. I mean, actually even as an experienced trader you’ve got instinct too often tell you the wrong thing to do but at least you’re aware that your gut instinct is often the wrong.
Adam: Exactly, yes. It’s one of these things you’ve got to learn to go against, basically. It just comes from experience, I think. I’m just having loser off to loser handle to play and start thinking and say, “Hang-on, maybe I shouldn’t be doing this.”
Walter: Yeah. Exactly
Adam: Well, many people they think about trading, many people give up before they learn.
Walter: Currently, now, you have some of your own systems that you’ve created and that you’re currently trading or that sort of babysitting these algos or is that not part of your trading or are you discretionary and sort of semi-automated or fully-automated systems.
Adam: It’s instinct because everyone is always amazed that someone like me, with my programming ability — and I can write EAs with my eyes closed actually but I don’t do more algorithmic trading.
The truth is, in all my years of programming, it is really hard coming out with stuff which is purely automated, rule-based, and makes money. I also think the truth is the best edges is to be found by having some simple setup which you could program up but then it’s the discretion is still when you use it, is that the edge that makes the money.
Let’s say, something like Jorge’s system which we have in the forum. It’s a very simple setup it is enough to program up the rules. There’s no doubt about when you’ve got 2Es setup and how he manages the trade are all programmed, all that up. I could write an EA, no problem.
The reason why he makes money on it and the reason why a complete, automated EA doesn’t make money is that he chooses carefully the market conditions and which setup he’s going take. The set-up looks a bit weird, I’m not going to take it. And then, he likes it, he takes it, and that is why he makes money.
I can’t get around this at the moment, the fact that it is very hard to find purely automated stuff which actually does make money. It is possible. There are folks out there who are doing and I have — through my work over years — come across certain outfits which do manage to do it.
There is a guy working at the moment. He has a purely automated genetic algorithm optimizing a thing and he makes 2-3% a month with less drawdown and grinded it up. And, there’s another guy incredibly complicated real time analysis of the US stock markets and dig all these stuff real time. His performance is phenomenal.
He makes 30-40% a year with drawdown about 2% which is insane. But, sadly, these examples are few and are far in between and it’s not easy to make something purely automated. To answer your question, I am actually more into discretionary trading.
In fact, at the moment, I am looking to a scalping approach come across which I actually found it suits my style and I can’t use make money using it. It’s not automated by any means.
Walter: Right. Do you think that — because I’ve often wondered this — do you think that one of the reasons why it’s difficult to program a profitable system for many traders and programmers is because there is something that the trader is looking at or analyzing?
It’s not really being articulated so therefore it doesn’t make it into the code. For example, like your Jorge when he trades this trend with the 2E systems, there’ll maybe something that he recognizes in the chart but is not something that he would tell you if he is looking at when you were to write down the rules of the system.
Adam: Yes, I think you’re right. The rules of the setup are obvious and you can write those down. The other thing which he is looking at, even if he told you, it’s the nature of how the bars are all hanging there together, whether there is a support resistance is.
How much of a stronger trend is, how choppy it is, all those other things or even if you articulate them — and I’m sure Jorge could. The difficulty is actually coding stuff up like that. That’s really hard. It’s very easy to code up something like a moving average, crossover, because it’s so obvious, it’s no debate about it, but is the trend strong enough?
You could possibly come up with some rules at all whether you’re support resistance is getting hard because picking out support and resistance is not an easy thing because we do it by either with the chart. It’s here but actually programming that up is hard.
It’s reading the market poll and all that kind of stuff is just so hard to come up with a purely ultimate objective way of doing it. It’s that sort of intuition we have coming with experience that way of reading the market. That’s difficult. If you could do that fully automated then it would be… The whole thing would be too easy.
Walter: Right. I like the idea of semi-automated. I mean, I guess I’ve done quite a bit of that over last couple of years. Sort of just will use discretion to determine which market I’m going trade in and then I just turn on the set-up for a system that I like to trade. But, I don’t know that I could articulate very well the rules that I’m using to essentially filter out different markets so I suppose that sort of fits.
Is there something that you would say as a trader and as a programmer and some with your experience that you’ve notice that when people go down this path, that they often make the same mistake or there’s a set of mistakes? For example, maybe they over optimize so they have too many parameters.
They have perhaps too many parameters or they over optimize or they don’t use like in-sampled/out-sampled data to confirm that what’s going on is going to work with an in-sampled data or they don’t use walk forward analysis. Is there something that you can say, “Look, definitely this is one thing or 2 things or 3 things that you should look at this when you’re building an automated system”?
Adam: Some of the things you have touched on, they are actually quite important. I would say that, in general, it’s looking at the biggest picture. The issue to deal with the fully-automated stuff — and it’s something that I touched on earlier — is that the goal post keep moving in that you can come up with something.
You optimize over the last couple of years or something like that and you think, “Yeah, this works really nicely and great lovely backtest equity curve”. And then, what happens is that the market just seems to shift flat in a way that it behaves in that sort of setup that works perfectly in the past suddenly doesn’t work anymore.
As a general rule of thumb, the shorter term of the stuff you’re doing, I think the more prone the market to this. I talked about paradigm shift. The market undergo paradigm shift like this and suddenly it wouldn’t pay this way. There’ll maybe some underlying fundamental which have little bit shift. There are people out there, expecting the whole time, analyzing the market.
They too, if they spotted the edge, they’ll be exploiting it too. The edge soon will disappear. So, bigger picture thing but going there into more detail. For sure, you need to have a lot of powerful testing and over fitting is so alluring. Add one more parameter suddenly, everything else looks better but you have to ask yourself whether it’s actually predicting anything or you are fitting to the backtest data.
It’s a very easy trap to fall into. It’s not easy coming up with the fully automated thing but I’m just going to go back on what you’ve touched on about it. Sometimes, switching on a EA and using — I mean I use and I have EA’s for managing trading stops, for example. Some setup I use as, if I’m in this decision where I can manage a trading stop, I’ll just drop an EA on it and I just walk away, letting it do its business.
So I think, as a great believer in automating the bits that can be automated, I also personally believe that it’s very hard to find something that’s fully-automated and feel like you need discretion overlay on top of it to know when to use your EA and when not to trade at all, that kind of thing.
Walter: Right, interesting. So, in your current trading, when you’re trading in your scalping system, is that something that you would do? You’re in a position that’s profitable, you might drop the EA on there and just let the trailing exit take that trade out or something like that. Is that your approach?
Adam: Yeah. I’ve used a couple of setups in those scalping thing. One of the things I’m still developing which is — but it’s almost like a big shadow on 5-minute chart — looking for big impulsive moves and you want to get in on those and have some kind of trailing exit.
Certainly, it’ll do for that particular type of set-up. No, I mean scalping, in general, is more than just grabbing a few pips when you’ve got it oten roll on trailing things but I do have a one particular set which is where this impulse trade where I will indeed drop a trailing exit on it and just leave it.
Walter: Right. And, so right now in your trading, what which timeframes are you focusing on. Do you focus on a 1-hour charts or the daily charts? I suppose it depends on what the markets are doing which ones you will spend time on but in a typical day or typical week what will you be looking at?
Adam: Yeah. I’m pretty into scalping at the moment and that is done on a 5-minute chart. I’ve been caught interest by a chat called David Franklin. He is a professional scalper, he’s making a nice living from it. He is very willing to share. He’s busy paying it forward.
He is being sort of helping out a number of people. I’ve always been drawn to the short term timeframe. It just suits my mentality. I like the effort of the instant feedback you get with short term stuff. I know that for many people they’re not suited to full timeframe.
I know you’ve mentioned how studies have shown that sort of trading on short term, 1-minute 5-minute bar charts, are like that addiction area in your brain and things like that, that’s certainly not going to be for everyone. Personally, I very much enjoy trading a 5-minute charts and find out more consistent doing that many of the things I have looked at.
Walter: Well, that’s fascinating. So that’s pretty much your preferred time frame, then?
Adam: Well, it is of the moment. So, I am just, exactly, I’m using it. It’s a 5-minute bar scalping approach. You just switch on the chart each day, you follow this 5-minute chart to look for those little setups and things like that, and that’s where you go.
Walter: Right, excellent. I am just curious. Do you have certain rules in terms of you can only have so many? Like, you have let’s say 2 losers per day then you’re out, that’s it for your day? Or, do you have sort of risk parameters that will get you out of a really poor day or something like that?
Adam: Yes. In fact, David Franklin himself — I mean I’ll just give you a bit of a background of what David does — is typically looking for 3 to 5 pips and some very small little scalps like that. Stop loss initially just to be quite large but something like 15 pips maybe.
If he’ll find himself on the wrong side of the trade, he’ll mostly bail out early. He has a rule that 3 losers of any size and is done for the day. He’ll stop because this sort of inverted risk/reward type of trading, a loser has such a good proportionate impact.
Often, he’s doing 30-50 trades a day sort of thing and he may, will have a loser at all. He’s been doing 5-6 years I think, he’s very consistent. Personally, if I find I’m going to have 2 losers, I’ll certainly stop for the day. One of the things I like about scalping is that the whole process is very accelerated.
In trading, we’ve got to learn about balancing the risk/reward against the hit rate and also it’s ultimately about our own psychology. It’s about what’s going on in our heads, which is getting in the way of making the profits. The thing I like about scalping, the whole process is very quick.
You can have 10 trades in a 2-hour session and you can have your losers. It’s staring you on the face, you’ve taken that emotional hit. A large part of what I’m doing while I’m trading, while I’m doing the scalping, I’m spending all my time monitoring my emotional state.
If I find myself out of filter then I will not take anymore trades because I just knew it, I can’t think straight. It’s a very sort of speeded up version of trading. You try out a 4-hour chart, you’ve taken a lost of 4 hours to recover before you can even think about the next trade. You see you might have 5 minutes. I’ll be bored then.
Walter: I will think that you’d have to fight against the revenge sort of thing. “I lost last time. I’m going to get you back. You got me 9 pips last time.”
Adam: The revenge is a real killer, the revenge trade. You really need to be objective about it. The thing is — there is an old time friend I know — as I said you have more than 4 hours to recover or whatever but it still hangs over you on the last trade as a loser. I’ve got to make that back somehow and that I know, personally is a great stumbling block for me, is that to do it revenge trading.
I’ve done your fantastic You Get What You Trade For in the Next Forex Now Forum. I know what my psychological buttons are that when I get pushed out, I don’t behave rationally. One of them happens to be having stuff taken away from me and it will extend from childhood.
It’s nothing particularly big that happened to me but I remember a few incidents that in Nursery School, first school, some boy taking something away from me. My parents would sometimes find me for doing something. I just have issues to deal with stuff being taken away from me so I’m very aware of this in terms of a revenge trade.
If I have a loser, I felt like it’s being taken away from me. I better make sure it’s not resonating with that past issue and making me trade irrationally. Trying to get it back, trying to fill that emotional hole which is you’ve got to be this passionate and objective about your trading. We’ve got to work our way through these things. There’s no getting around this.
If you want to be successful as a trader, you have to work through all the stuff so it’s no longer influenced. The thing I love about scalping is that whole process of working through it and seeing these things happens on a much fall of a time scale. You make 2 trades a week on your 4-hour charts. I’m getting 10, 20 trades in 2-hours so you’ve got plenty of opportunities to address this issue.
Walter: Yeah, that’s interesting. Do you find that, when you start making money on the quick trades scalping trades, do you find that your broker starts to slip you and funny business like that? Do you see that happening?
Adam: Not personally because rule 1 for me of scalping is you have to use an ECM broker, a proper ECM broker. So, it’s got to be a broker platform which has multiple contributors with wholetime credibility and the brokers are just taking commission. Otherwise, there’s inherent conflict of interest between your broker and you.
As you say, no broker is going to use… If the broker is taking the other side of your trade when you’re starting to make money, he’s going to do something about it. It’s just logic who you’d taking there so you have to put this kind of trading, you have to use an ECM broker. Fortunately, there’s quite a figure one out there. That’s rule 1: don’t even attempt scalping unless you’ve got that in my book.
Walter: That’s great advice. It’s interesting because you’re by far the expert at drawing up automated trading systems and yet here we are talking about discretionary scalping.
Adam: All I know, people are always amazed at this. How come someone like me so drawn to this discretionary part of it? I’d like, I went with some underlying psychological thing behind it, I don’t know. There might be but, certainly, there’s nothing wrong in trying in terms of purely mechanical stuff.
I’m thinking on the fly here but actually, now that you’ve mentioned it, there are purely mechanical things that you can get to work. There’s things like hoping range breakouts stuff. I think this guy in the Naked Forex Now Forum had some Asian range breakout thing and you backtest those things, they do make money.
But, the problem with all purely mechanical trading strategy is that they do have periods when they don’t work very well. They have those long periods of drawdown and I think you’ve got to either decide if you’ll just be a purely mechanical trader, in which case you just going to have to sit through those drawdowns, or you’re going to have to go down the discretionary route.
Personally, I’ve always found it uncomfortable sitting through — I mean, everyone doesn’t like drawdown — I’ve always stand uncomfortable sitting through these drawdowns. I’ve always been looking for this Holy Grail, this little something that’s consistently making money so you don’t have to endure this. Personally, I’ve been drawn to this discretionary side is to try and get out from that sort of the drawdown you get on the mechanical scale.
Walter: Right. What about reversion to the mean style of trading? I mean, typically, they still have drawdowns but they can have fairly nice looking equity curves, right?
Adam: There are outfits who do — algos outfit — who do these reversion to the mean stuff on a bust scale but even they struggle on it. There’s something very alluring, psychologically, about trading reversion to the mean. I think it’s what the Turtle guys call it, they call the “counter trend”.
It’s very alluring but I don’t think that’s where the edge is if you test these things. I’ve yet to see a decent counter-trend system which actually holds up. I mean, there’s all these trading nonsense out there which is very alluring because it seems like it almost never loses money until the one time on the market just keeps on going and then wiped out your account.
That’s not the best I am willing to take in. I’m willing most of the time but risking up, walking up my account. To say that good trading stuff is a normal starter and there are some people who successfully trade a counter-trend.
I would say that, just like in all trading, it’s the discretionary overlay which makes the difference. People are carefully picking their levels when where they’re doing this than managing their trade right. They know from experience how much they can push it and all that kind of thing.
I’m actually certain that can be made to work just as some trend following path that can be made to work. It’s not necessarily any easier an option. I think that there’s no doubt that there are underlying trend vices in the market because that’s basically what our long term hedge fund stuff relied on at the end.
Those vices are there and you can look at the distribution of the market returns. They all have this so-called fat tails which basically means the market will tend to move more than you might have think it would if it’s in random.
Darren is another person in the Naked Forex Now Forum, a very successful trader. He’s always been talking about the levy flight distribution, how markets are like this. Well, this occasionally move further for longer than you think and all these tends to tell you that you are really trading on the trend side rather than on the counter-trend side. That is for me, anyway.
I’ll say one of the psychological biases that something will be drawn to the counter trend and no doubt some people can make it work but it’s certainly not for me, personally.
Walter: Yeah, exactly. If you know what sort of distributions you’re going to see when you are looking at trade results in the markets it makes a lot of sense to use a trailing exit. At least then you’re going to know that you’ve a really big huge winner who comes into equation and bumps up your average winner, right? That is basically what you’re looking for.
Adam: Absolutely, yes. Those occasional big winners can make all the difference. I mean, there are some strategies will just rely entirely on these occasional huge winners to make them profitable.
Walter: Yeah. I love the trailing exits. You can even do without a trailing exit, you can have an aggressive target but the point is, if you’ll use a trailing exit, in theory you are allowing the market to go as far as it wants to go but it can be difficult to trade it this way.
Adam: Of course psychologically they’re quite difficult to trade because so many times you’re still on a modest profit only to end up giving it all back and you end up with nothing. So it’s quite difficult, but if you’re prepared to sit on your hands then occasionally you’ll catch a huge winner which hopefully makes up for all those times when you gave it all back so that’s the psychological trade-off really.
Walter: Yes it is difficult, I agree but there are a couple of interesting points. Listening to what you have to say here Adam, I know that it is easier for many traders to trade a higher win rate. I’ve noticed, I’ve tested traders around the world and most traders are focusing on this.
It’s easier psychologically to do that but that is the kind of allure of scalping, isn’t it? You really have to be careful in your scalping because one, you’ve got to keep your win rate higher. Otherwise, you’ll get into trouble in terms of making money. Secondly, you have to pick your broker very carefully. As you know because as soon as your broker knows what you are up to, it’s very easy for them if they are market maker to steal from you and make it difficult for you to make money. In fact, a lot of ECN’s that are out there, they claim to be ECNs but they are not true ECN. People do not realize this and they think that they’re going to go and see “Oh, it’s an ECN. Okay, great. I’ll go with this broker” but it is not that simple.
Can you talk a little bit about that? About how you discerned between a broker that you’re comfortable with scalping versus those that you should definitely stay away from if you’re just looking to make a few pips here and there with your trading. In other words, how do we spot a true ECN?
Adam: How to spot an ECN, yes that’s a good question Walter. Well, I think with some brokers you can actually look at the depth of market and it shows you the name of the bank providing the liquidity for each price. I know if you use Interactive Broker for example then you can actually see all the contributing banks and there’s all the big names so you know that it’s definitely an ECN.
If you’re using something like Metatrader though on the other hand then you don’t get that transparency, you can’t actually see who’s providing the quotes. You can sort of tell by looking at how the prices jump around but also how tight they are but probably the simplest way of picking out a broker is actually by word of mouth.
That’s one of the great things about something like the forum, the Naked Forex Now forum, there’s plenty of people around to ask, they’ll tell you what brokers they recommend, it’s one of those regular topics that comes up all the time. There are a few ECN brokers that I recommend and if you look in the forum then you can see which ones they are. So that’s probably the way that I’d go about choosing a broker.
Walter: Yeah, exactly. Thank you so much Adam for spending time with us here today. I am looking forward to hearing your presentation at the conference. Somebody with such a wealth of experience working for hedge funds, trading yourself and having that background in Psychology, I think it’s going to be a great presentation for those who are able to attend and I really appreciate your time. Thanks so much and we’ll see you at the conference.
Adam: Okay, bye.
Podcast: Play in new window | Download
Subscribe: Apple Podcasts | Android | RSS