Psychologically Robust Systems

Discussions about the testing and simulation of mechanical trading systems using historical data and other methods. Trading Blox Customers should post Trading Blox specific questions in the Customer Support forum.
Post Reply
Forum Mgmnt
Roundtable Knight
Roundtable Knight
Posts: 1842
Joined: Tue Apr 15, 2003 11:02 am
Contact:

Psychologically Robust Systems

Post by Forum Mgmnt »

In another topic I wrote:In general, I also like to have some rationale for why I believe a system works. I want the source of the profit to be based on something that doesn't change. Systems that rely on human psychology and emotions tend to hold up very well because humans don't change very much, even when they really want to.
Then a bit later Marc (a.k.a. DutchTrader) wrote:You made a nice point about robustness i.e. the psychological part of all participants of the markets. Can you tell me/us more about your view of this topic. I think it is very interesting!
Yes, it is an interesting topic.

Consider a long-term trend following system, there are three important psychological considerations that give me reason to believe that long-term trend-following systems are psychologically robust:
  1. Poor psychological memory - Humans tend to overweight the recent past in their judgements. This means that most people have a hard time buying something that is at it's highs, or selling something at its lows. There are a lot of behavioral finance articles on this sort of thing.
  2. Need to be right - Humans tend to want to be right more than wrong. They don't like doing things that have a low win/loss ratio. Long-term trend following generally has a win/loss ratio of less than 50%.
  3. Aversion to Pain - Humans can't stand pain. The long periods of drawdowns of many months or sometimes a year or more, make it very difficult for people to follow systems of this sort.
All of this is predicated on my belief that it is easier to make money when most people are not willing to enter a position. The more people are predisposed to take a position, the choppier the trend that results, since there will be weak money panicking when there is a retracement. This panicking can cause money management stops to get hit resulting in losses even during trends. Conversely, when there are few people taking positions and the price moves without their participation, there is less energy for a retracement, and moves tend to be much smoother.

This is also the reason that I believe certain markets are easier to make money in. If the speculative trading has a significant effect on the price, the markets are choppier, if the trading is mostly non-speculative, the markets tend to have smoother trends.

Compare Gold or Pork Bellies to Eurodollars and the Japanese Yen. In the former case, speculation, especially by neophytes is a large percentage of the reason for price movements. In the later case, you mostly have global economic issues at the governmental level that affect prices, most trading is based on hedging or market making rather than positional.
blueberrycake
Roundtable Knight
Roundtable Knight
Posts: 125
Joined: Mon Apr 21, 2003 11:04 pm
Location: California

Re: Psychologically Robust Systems

Post by blueberrycake »

Forum Mgmnt wrote: Consider a long-term trend following system, there are three important psychological considerations that give me reason to believe that long-term trend-following systems are psychologically robust:
This may well be true, but I think its important to consider the evolution of computerized trading. In managed futures, there has been an tremendous growth of assets under management, with many of the funds being managed on trend following principles. While I don't have exact numbers available, I would not be surprised if more than 50% of the off-the-floor speculative trading was done with the help of mechanical systems, for which there is no such thing as a psychological consideration.

I am not disputing that trend-following works, since testing clearly shows that it works quite well. I am skeptical though that the reason it works is for the psychological reasons that you had described.

-bbc
Forum Mgmnt
Roundtable Knight
Roundtable Knight
Posts: 1842
Joined: Tue Apr 15, 2003 11:02 am
Contact:

Post by Forum Mgmnt »

bbc wrote:While I don't have exact numbers available, I would not be surprised if more than 50% of the off-the-floor speculative trading was done with the help of mechanical systems, for which there is no such thing as a psychological consideration.
On the part of the traders, perhaps, however I wouldn't discount the effect of the drawdowns on the investor's psychology. It is very, very hard for a long-term trend follower to raise money today. The money runs away faster from investors withdrawing money than it does from trading losses. Someone with a $500 million under management who loses 30% can expect $300 million to leave in addition to the $150 million that was lost from trading, resulting in a 90% drawdown in assets under management.
I am skeptical though that the reason it works is for the psychological reasons that you had described.
I was simply making the case for long-term trend following being what I call Psychologically Robust. It is only one factor, there are many others that fall outide the area of psychological robustness.

If it were easy to trade money this way, there would be little opporunity to make money.

Why do you think that a technique that has consistently made money over the last 30 to 40 years is still viable if not for psychological factors?

Contrast this with the typical arbitrage opportunity where the influx of players quickly renders an arbitrage obsolete after as little as a year or two in typical cases.

I'm a big fan of anything that works. I trade and research many other types of systems besides long-term trend following. Any systematic concept with the right risk-reward characteristics works for me.
Bondtrader
Roundtable Knight
Roundtable Knight
Posts: 101
Joined: Mon Apr 28, 2003 11:07 am
Location: Zimbabwe

Post by Bondtrader »

The CFTC website has a "Commitment of Traders" report from 20 May 2003. If I read it correctly, hedgers are called "Commercials" and speculators are called "Non-Commercials".

Eurodollars: 53% of all long positions and 80% of all short positions, belong to hedgers (non-speculators).

Pork Bellies: 24% of all long positions and 12% of all short positions, belong to hedgers (non-speculators).

The rest belong to commodity pools, managed futures funds, big swinging position traders, and little punters like me and you.
damian
Roundtable Knight
Roundtable Knight
Posts: 815
Joined: Tue Apr 15, 2003 8:43 pm
Location: dusseldorf

Post by damian »

Forum Mgmnt, this is a great topic.

I agree with your 3 propositions. At first I didn't see the cause and effect. Then I thought about your statement : "All of this is predicated on my belief that it is easier to make money when most people are not willing to enter a position".

I understand you to be saying, in very brief paraphrase, that trend following works becuase less people are willing to take a position when a trend starts. The reasons less people are willing are explained in 1, 2 & 3.

I may be wrong in that interpretation, please tell me if so as my following ponderance is based on my interpretation in the above paragraph.

How do 1, 2 & 3 cause trends? I understand that they make for good smooth trends, but I do not see how they directly cause the trend.

Essentially, my view holds that trend following will work so long as:
  • a) trends exist,
  • b) they are relatively smooth trends owing to your propositions 1 thru 3.
MODERATOR'S NOTE: Changed the BBCode to properly format the list. NOTE: See faq.php?mode=bbcode#6 for the correct way to format lists.

Have I misinterpreted your theory? If not, I would like to continue to discuss what psychological reasons cause (a) to exist in the first place. Are they hope, fear, greed? If so, I can read hope, fear and greed into your propositions... so am I just circling back on myslef and addressing the same issue with a different accent?

Is a long trend always driven by an element of "public mania and popular delussion"? Or is it the absence of which cause a pre-existing trend to be trend follower friendly?

damian
Forum Mgmnt
Roundtable Knight
Roundtable Knight
Posts: 1842
Joined: Tue Apr 15, 2003 11:02 am
Contact:

Post by Forum Mgmnt »

you wrote:How do 1, 2 & 3 cause trends? I understand that they make for good smooth trends, but I do not see how they directly cause the trend.
In short, they don't. The trends are caused by other things, most often some actual fundamental reason based on supply and demand or bargaining power considerations.

By my read, your interpretation is correct. 1, 2 & 3 make it possible to make money easily from trends which as you alluded is what makes trend following work. So my 1, 2 & 3 cause the B portion of your requisite conditions for successful trend following, not the A condition.

We must have A (trends) and B (smoothness) to make good money. A without B can make some money or even lose money. B without A is not good either.

Hope tends to make people keep positions they should have exited.

Greed makes people take positions they shouldn't (i.e. long after the trend has started to run) and keep positions when the should exit after the trend ends.

Fear keeps them from getting in early like they should and from getting out with a loss when they should after they do enter.

Hope, Fear, and Greed can indeed add momentum to a trend that has already developed but rarely do they cause trends of any real substance.
bagherra
Senior Member
Senior Member
Posts: 33
Joined: Fri Apr 25, 2003 9:03 am
Location: Charlotte, NC

Psychologically Robust Systems

Post by bagherra »

There are other methods, such as value investing and long/short equity hedge funds, that work. However, the public finds it difficult to keep money in a system that underperforms over long periods.

Most value investors underperform the market between 30% to 40% of the time, even though their results, taken on the whole, beat the market by 5 to 10 points or more annualized. (So if the market returned 10% on average annually, value investors got 15% to 20%).

Long/short equity hedge funds beat the market by approximately 7 points annually [after fees] during the last twelve years. Most of the excess return, however, has occurred during the bear market of the last three years. (Long/short equity funds lagged the market by approximately 2 points per year on the way up).

Thanks to the recent outperformance, long/short hedge funds are now popping up 'like lemonade stands'. Most of these investors won't be able to hang on once the stock market begins another climb, because they'll be chasing performance, which is what got them into hedge funds in the first place.

The moral: some long term methods of beating the market work, such as trend following, value investing, and long/short equity, but people can't stick with them for psychological reasons. The trick is to change our investors' psychology, which is very, very hard.
Last edited by bagherra on Sat May 31, 2003 10:34 am, edited 1 time in total.
Sir G
Moderator
Moderator
Posts: 243
Joined: Wed Apr 16, 2003 12:21 am
Location: Salt Lake City, Utah

Post by Sir G »

Great topic, c.f.-

This is the kind of stuff I love. I’m coming from the short-term world so I’ll add in my 2 cents from that location.

I agree with you 100% except for the purpose of those thoughts. I think those thoughts are perceptual filters that we use to give us confidence in what we do. It’s a psychological edge that we can apply to ourselves. The numbers that our computers run can nurture our intellectual edge, but some of us, myself included, need to take our systems beyond the math and the logic that gave birth to that math and move it into a solid foundation of why it will work in the future. In trading we just don’t have to believe in ourselves but also our systems and their results in both the historical and real-time worlds.

In my short term trading, I’m sure on any given day I have traded with and against you, Richard Dennis, Toby, Tudor Jones and most everyone in this forum who has traded. You’ve taken my paper and we have competed for the same paper. Anyone who has traded can make this claim. I say this to highlight all the overlapping that our systems have.

I think your 1-2-3 can apply to all styles of trading.

1. Poor psychological memory:

If your time frame is just a few days and if yesterday closed strong, the feeling one may get when a buy order appears maybe the same feeling a longterm trader buying the 50 day BO gets.

2. Need to be right

I think this loops back to #1. Very few can keep an accurate score of what was, so the behavior, desires and focus become the next trade, not an accumulation of past trades.*

3. Aversion to Pain

This is an issue that is shared by all participants of trading. Unfortunately draw downs are not the exclusive domain of LT traders.

In general, in my eyes there is no difference between a 3R trade a long term trader will capture and a 3R trade a short term trader would capture. Even the risk adjusted returns….Everything is relative. Once things get normalized, the appearance of any exclusivity from one time frame to another becomes irrelevant.


Gordon

*The discretionary trader and the mechanical trader will have different views but the potential for the same outcome. Can a system trader fool himself enough by being so callous in his testing to fool himself by optimizing.. sure. Can the discretionary trader grab for quick profits, sure.
Chuck B
Roundtable Knight
Roundtable Knight
Posts: 481
Joined: Thu Apr 17, 2003 6:34 am

Post by Chuck B »

Once things get normalized, the appearance of any exclusivity from one time frame to another becomes irrelevant.
Agreed. The only difference perhaps is the role that frequency of trade entries plays in one's psyche, how often a trade opportunity presents itself. It is possible that a short-term trader will experience a far greater diversity of outcomes within the same length of time than a long term trader. It may take 5 years of portfolio trendfollowing to experience that mother of a string of losing trades or that massive run-up that astounds one's sensibilities in its size. Hey, with a one-minute bar trendfollowing system in the SP, you could experience this in a couple of weeks :lol: .
Kiwi
Roundtable Knight
Roundtable Knight
Posts: 513
Joined: Wed Apr 16, 2003 1:18 am
Location: Nowhere near

Post by Kiwi »

Not wanting to let this discussion stay so pure and edifying :wink:

I was just reading an excellent article recommended in another thread about Hostetter. In it was a quote about one of this groups great psychological props :twisted:
At Hostetter's urging, Weymar assigned Vannerson to set up a technical system known as "trend following." In its simplest form, trend following instructs a speculator to buy a commodity whenever the price goes up, and sell it when the price goes down. An investor has to get in or out of the commodity every time the price reverses, so that he's constantly whipsawed by small losses. The goal, of course, is to outstrip those losses by catching a long upward or downward move.

Vannerson aimed at a more sophisticated trend-following system. To avoid being whipsawed, he wanted to figure out the pattern in which a price starts up or down at the beginning of a real trend. He computerized the daily prices of 15 commodities over at least ten years and found that each commodity has its own "personality." In a volatile commodity such as pork bellies, a sharp drop in price in a single day signaled a decline of at least two weeks, while goods such as corn began a downward sweep by slipping gradually over several days.
I used to believe that all commodities should be traded with one set of parameters. Then I started discretionary trading pouring over years of data and trading my theories forward one day at a time. I'm afraid that gold is not the same as wheat or yen or (heavens forbid) sugar. Surely this is driven by some of the combination of psychological factors discussed here. So why not adjust your trend following system to the commodity being traded?

John
Last edited by Kiwi on Fri May 30, 2003 8:34 pm, edited 1 time in total.
blueberrycake
Roundtable Knight
Roundtable Knight
Posts: 125
Joined: Mon Apr 21, 2003 11:04 pm
Location: California

Post by blueberrycake »

Kiwi wrote:So why not adjust your trend following system to the commodity being traded?
This is a really good question, that has been frustrating me for quite some time. For example, if we look at I-Master, we see a system, that works only on index futures, and worse yet, it only works on post 1996 data. Nevertheless it has held up quite well since release.

At the same time, I was recently reading one of Toby Crabel's "Due Diligence" questionnaires, where he asserts that he daytrades all his markets (index futures, interest rates, energies, currencies, etc.) using the exact same system with the exact same parameters.

I suspect that the only way to know for sure is to backtest the actual methodology for "discovering" the system, using both market specific and generalized rule sets. Sure seems like a LOT of work though =)

-bbc
MCT
Roundtable Knight
Roundtable Knight
Posts: 102
Joined: Fri May 16, 2003 7:27 pm

Post by MCT »

In short, they don't. The trends are caused by other things, most often some actual fundamental reason based on supply and demand or bargaining power considerations.
I agree with c.f.’ comments …

human psychology is fractal in nature …as c.f. pointed out:
Hope, Fear, and Greed can indeed add momentum to a trend that has already developed but rarely do they cause trends of any real substance.
I believe the real causes of sustainable long-term trends are “ wrappings that seem to occur in society’s rationality.â€
Last edited by MCT on Wed Sep 17, 2003 7:49 pm, edited 1 time in total.
Dutchtrader
Roundtable Fellow
Roundtable Fellow
Posts: 58
Joined: Wed Apr 30, 2003 4:35 pm
Location: Netherlands

Thanks..

Post by Dutchtrader »

Thanks c.f. for the great reply and of course all other members who contributed to this topic. I learned a lot and it gave me some ( creative )inspiration for system development.

Marc
Post Reply