Forums  > Trading  > Losing algo-traders support group  
     
Page 1 of 2Goto to page: [1], 2 Next
Display using:  

Energetic
Forum Captain

Total Posts: 1506
Joined: Jun 2004
 
Posted: 2019-10-16 21:20
Sorry for the subject line - could not resist.

As someone here recommended, I've read "The Market Wizards" book. One of the things that I took away was that several wizards said that, for the purposes of risk control, they take the chips off the table when they lose certain percentage of money and then gradually return back to the market as their confidence restored.

I thought it's a reasonable thing to automate so I tried to apply several obvious ideas (i.e. reducing the bet size as the drawdown deepened, at various speeds) to my strategy ... and failed miserably. In the best version, the profitability and ratios declined only modestly while the max drawdown became a just bit shallower. In all other versions, I lost performance without improving drawdowns at all. The reason is that the strategy tends to recover fairly quickly on its own so by the time I was beginning to pull the money away it was typically just about to rebound.

Has anyone tried to do something like that? Any success?

For every complex problem there is an answer that is clear, simple and wrong. - H. L. Mencken

tabris


Total Posts: 1264
Joined: Feb 2005
 
Posted: 2019-10-17 01:07
"As someone here recommended, I've read "The Market Wizards" book. One of the things that I took away was that several wizards said that, for the purposes of risk control, they take the chips off the table when they lose certain percentage of money and then gradually return back to the market as their confidence restored."

My guess that this worked for them is that they have a decently high autocorrelation on their returns or their style/strategy has an inherent bias towards momentum.

"I thought it's a reasonable thing to automate so I tried to apply several obvious ideas (i.e. reducing the bet size as the drawdown deepened, at various speeds) to my strategy ... and failed miserably. In the best version, the profitability and ratios declined only modestly while the max drawdown became a just bit shallower. In all other versions, I lost performance without improving drawdowns at all. The reason is that the strategy tends to recover fairly quickly on its own so by the time I was beginning to pull the money away it was typically just about to rebound."

My guesses at why this did not work for you is... imagine you play blackjack where you have an edge. But you reduce your bet size when you lose money not when your edge goes lower. This will most likely result in what you described. So, the answer/solution to that would be, instead of looking at drawdown... you might want to reduce size on edge decay which is probably a harder measure. Second guess as to why this might happen is your returns might actually be negatively autocorrelated or your strategy trades some type of mean reversion. In this case, the optimal bet size is actually doubling down as edge has actually increased in your favor. By reducing bet size on drawdown, you are actually betting less when your edge has increased which will almost always end in ruin.

Dilbert: Why does it seem as though I am the only honest guy on earth? Dogbert: Your type tends not to reproduce.

london


Total Posts: 326
Joined: Apr 2005
 
Posted: 2019-10-17 02:51
A few random musings:

Reducing risk after a run of poor performance might be a good thing or bad thing: depends on the strategy (or the trader!)

Good: when the strategy has stopped working; the world has changed; the edge you were capturing has decayed (either temporarily or permanently); then deploying less capital is good if the expected return is now negative.

Bad: if the edge is still there and just going through a temporary drawdown. The expected strategy return is still long run positive, but putting less capital behind each trade, it takes longer to dig out the hole.

So how do you know which is which? That’s the hard part!
Ongoing monitoring and having a view of when your strategy is not performing as expected.
1. Look at historical strategy attributes - per Tabris - autocorrelation; and also distribution of P&L per trade and per period, number of trades, etc, etc. Have a view on how different your live strategy attributes need to get before you think something smells funny. I’m a big fan of statistical process control and Deming. It’s not flashy or sexy but I want to know something might be different _before_ I’ve collected a long sequence of losses.
Its a small data problem- I want to know something is different before I have big data (and big losses and drawdowns).

2. Under what conditions has your strategy done well historically?; under what conditions has your strategy done poorly historically? And - critically - what conditions have not been observed historically?
I’m a big fan of knowing the range of conditions a strategy has been tested under to then understand where it has _not_ been tested. This is the true unknown and where to reduce the chips. Or at least pay closer attention.

3. Keep in mind you’ve never seen a strategy largest drawdown:a maximum (by definition) can only get bigger ;)

4. The other time when cutting risk on losses is a _very_ good idea: for discretionary traders.
Have a fixed loss per trade, per hour, per day, per week or whatever. Hard stops. No more trades.
If one gets breached: close positions, stand up, walk away. Don’t come back till the next period.
When trades are being placed discretionary (rather than automated); this avoids the human over-confidence and emotional temptation of trying to “trade out” of losses with increasing larger and larger bets.
Eventually, that ends only one way and it’s not pretty.

I suspect discretionary trading is not your context.
But maybe this is the context of Market Wizards? I’ve never actually read it.


So if reducing risk on losses is good or bad, depends on: the strategy in question; how discretionary the trader is and the current environment versus the historical environment when the strategy was developed.


ronin


Total Posts: 514
Joined: May 2006
 
Posted: 2019-10-17 10:02
Yes, welcome to trading 101.

Stop losses don't help, and they will ruin any strategy. The only way to keep them from ruining your strategy is make the stops so wide that they are meaningless.

In the real world, you are running the strategy because it has certain risk characteristic. A drawdown may be just the cost of doing business, or it may mean that the strategy no longer has the risk characteristics you want. Or never had them in the first place.

That is what you analyse, and that will then drive your decision what do do with the strategy, not the depth of the drawdown per se.

"There is a SIX am?" -- Arthur

Energetic
Forum Captain

Total Posts: 1506
Joined: Jun 2004
 
Posted: 2019-10-17 14:00
@ronin:

In other words, the wizards are actually rookies who didn't know what they're doing, never mind the mind-boggling performance results? Thanks, that helps a lot.

BTW, unlike (some of) them I didn't implement a stop-loss. My size functions were continuous.

Edit: Now it comes back to me that some of the wizards took an equally disdainful view on stop losses. But others have not. That was my other takeaway: there are many ways to skin the cat.

For every complex problem there is an answer that is clear, simple and wrong. - H. L. Mencken

Energetic
Forum Captain

Total Posts: 1506
Joined: Jun 2004
 
Posted: 2019-10-17 14:03
@tabris, thanks - that's a reasonable diagnosis.

For every complex problem there is an answer that is clear, simple and wrong. - H. L. Mencken

Energetic
Forum Captain

Total Posts: 1506
Joined: Jun 2004
 
Posted: 2019-10-17 14:36
@london

Thanks for taking time!

Of course, I am always concerned whether the edge is gone. One possible view on drawdowns is that the edge goes away temporarily and then comes back. Or it's just a string of bad luck. It does, by the end of the day, come down to autocorrelation which I doubt I can predict with any accuracy.

1. I do have my own version of ongoing monitoring. I described it in this thread. General idea at the top of the thread and some details a little later in the comments.

2. Oh, I obsessed over this question a lot but I don't have the answer. Sometimes it looks like the signal temporarily disappears even though the market seems to be unremarkable. Actually, in terms of returns (not DDs) on a vol trading strategy a bad time is now. And I think I know the reason: someone is posting to Twitter too much ;)

3. Yes, it's good line

4. Correct, I'm not discretionary. My "product" is an algo.

For every complex problem there is an answer that is clear, simple and wrong. - H. L. Mencken

gaj


Total Posts: 54
Joined: Apr 2018
 
Posted: 2019-10-17 14:39
I think the pure quant approach would be Bayesian. You start by predicting the return distribution P(X). The sizing will naturally be a function of P(X). As new data comes in, automatically adjust the distribution to P(X | new data). This automatically readjusts your model if the new data supports or contradicts your hypothesis.

This requires modelling the full distribution P(X, data) beforehand though, which may not be so straightforward.

gmetric_Flow


Total Posts: 28
Joined: Oct 2016
 
Posted: 2019-10-17 14:45
Take a look at chapter 8 on Money Management in Filthy's book- you might find the section on Alternatives to Kelly fairly interesting. Alternatively, your favorite gambling book should discuss betting progression and of course the different schemes and how they fit into different utility curves.

JTDerp


Total Posts: 58
Joined: Nov 2013
 
Posted: 2019-10-17 15:05
(removed)

"How dreadful...to be caught up in a game and have no idea of the rules." - C.S.

Energetic
Forum Captain

Total Posts: 1506
Joined: Jun 2004
 
Posted: 2019-10-17 15:23
@gaj

This requires modelling the full distribution P(X, data) beforehand though, which may not be so straightforward.

That's to put it mildly. I'm predicting P(X|benchmark). This helps me to monitor performance but doesn't help with sizing b/c this measure is not forward-looking.

For every complex problem there is an answer that is clear, simple and wrong. - H. L. Mencken

Energetic
Forum Captain

Total Posts: 1506
Joined: Jun 2004
 
Posted: 2019-10-17 15:30
@Flow

Thanks. I'm staring at my bookshelf as we speak and the book is not there. I'll have to remember who borrowed it.

I looked at a couple of papers where people tried to apply Kelly to this problem and didn't find them particularly useful.

For every complex problem there is an answer that is clear, simple and wrong. - H. L. Mencken

riskPremium


Total Posts: 21
Joined: Nov 2018
 
Posted: 2019-10-18 12:29
@ronin,

"Stop losses don't help, and they will ruin any strategy. The only way to keep them from ruining your strategy is make the stops so wide that they are meaningless."

I beg to differ here. Whether stops ruin your strategy depends on whether there is positive return autocorrelation observed on the same scale as your stops.

nikol


Total Posts: 853
Joined: Jun 2005
 
Posted: 2019-10-18 19:48
I see, "XYZ Market Wizards" are serial blockbusters. Which one do you read?

Yesterday news might make you better - 2 years of DD and now up again.
https://news.efinancialcareers.com/uk-en/3002412/caxton-associates-europe-loss

Craig


Total Posts: 47
Joined: Oct 2008
 
Posted: 2019-10-18 22:12
@riskPremium
Can you expand on what you mean by "scale" of auto-correlation?


riskPremium


Total Posts: 21
Joined: Nov 2018
 
Posted: 2019-10-19 14:03
@craig

by scale i mean time frame. For example if you observe price momentum on hourly data, it might actually help if you place stops on your hourly strategy. Of course stops incur extra execution cost etc

Energetic
Forum Captain

Total Posts: 1506
Joined: Jun 2004
 
Posted: 2019-10-19 20:06
I believe I've read the sequel.

For every complex problem there is an answer that is clear, simple and wrong. - H. L. Mencken

nikol


Total Posts: 853
Joined: Jun 2005
 
Posted: 2019-10-20 11:16
Follow same path - reading.
Chap.8 tells about RM, stops vs options. But it talks also about techniques to isolate your mind from biased thinking, like "set exit before entry". Clearly this one cannot be implemented in algo trading.
I particularly like "the trader's dilemma" solution where procrastination becomes a sin.

xfd


Total Posts: 13
Joined: Mar 2008
 
Posted: 2019-10-20 14:32
Just a thought ... you say "my strategy" ... presumably singular. If the strategy is known ex ante to have favorable performance over its full history (guessing this is true if it's still running) that imposes some reversion of bad performance. Else it probably wouldn't still be trading. I wonder if you're applying the idea to a strategy that belongs to the subset of strats for which the method will be least helpful?

If the wizards' comments were largely about managing live strategies, any of which can break at any time, you might add some simulated strategies that work for x% of the history and fail thereafter. Then see if applying the idea helps portfolio performance.

ronin


Total Posts: 514
Joined: May 2006
 
Posted: 2019-10-21 08:49
> I beg to differ here. Whether stops ruin your strategy depends on whether there is positive return autocorrelation observed on the same scale as your stops.

You say "positive return autocorrelation" like it's a good thing. It's not.

So your point is that stops turn a fat-tailed strategy into an inversely skewed strategy. With worse performance than the fat-tailed strategy.

I mean, great. Whoop dee doo. Would you trade either of those?

No, I'll stick with my point. Avoid stops. And soft stops. Take the strategy off if it doesn't have the requied risk profile.

"There is a SIX am?" -- Arthur

Energetic
Forum Captain

Total Posts: 1506
Joined: Jun 2004
 
Posted: 2019-10-21 14:15
> I wonder if you're applying the idea to a strategy that belongs to the subset of strats for which the method will be least helpful?

Maybe but it wasn't obvious in advance of trying this. Just because the strategy is known to eventually recover doesn't mean that there is no way to modify sizing to mitigate drawdowns w/o sacrificing performance. And, just because I couldn't figure it out doesn't mean it's not possible.

For every complex problem there is an answer that is clear, simple and wrong. - H. L. Mencken

sharpe_machine


Total Posts: 32
Joined: Feb 2018
 
Posted: 2019-10-21 17:34
@ronin

> Stop losses don't help, and they will ruin any strategy.
> You say "positive return autocorrelation" like it's a good thing. It's not.

While I have seen it myself in the real world, could you please give us some sort of mathematical explanation of this?

Thanks.

ronin


Total Posts: 514
Joined: May 2006
 
Posted: 2019-10-21 18:19
> some sort of mathematical explanation of this?

Well, I don't know how mathematical you want to get.

In a nuthsell, stop loss on a long position makes you sell low, buy high. I.e., you are paying gamma. Short convexity. Every day you have a negative contribution to the pnl from paying gamma. It is a negative bias that you have to overcome through your superb stock selection.

Contrast that to buying low, selling high - here you make gamma. Every day you generate a bit of positive pnl, but every once in a while you are run over when your "buying low" becomes "catching a falling knife". But then, this is something you can diversify away. Each single-stock-falling-knife is worth only 1/n of your portfolio. You are still making a bit of gamma on each of them every day. So your upside is n/n*gamma, your downside is 1/n.

The risk that remains is a system-wide falling knife, which is the only thing you can't diversify. That's n/n. But that only really happens every ten years or so - with a big enough basket, you can pick up enough single stock gamma during those ten years to keep your expected pnl positive.

On the other hand, long gamma can't be diversified away. There is no mathematical way to generate positive pnl expectation - you have to rely on your stock picking ability. That's orthogonal to mathematics.

> "positive return autocorrelation"

Positive autocorrelation generates fat tails - kurtosis. You never have just one positive or negative return - if you have one, you have several. I.e., fat tails. Which is a bad thing in trading strategies. If I wanted risk in the wings, I'd buy catastrophe bonds.

But then you overlay a stop loss on your fat tailed strategy. So you make negative runs shorter, but leave positive runs long. In other words, negative skew. You are losing a bit of money every day, but counting on positive runs to get you back to positive. What happens if you miss one positive run every once in a while? Or catch it too late? Or keep it too long? Antifragile it ain't.

But then, yours isn't just any inversely skewed strategy. Your inverse skew is engineered by overlaying a stop loss. Which has negative gamma. See above.

So you have a massive negative bias hill to climb just to break even.

At one point you just go "no, this is just going nowhere."

"There is a SIX am?" -- Arthur

nikol


Total Posts: 853
Joined: Jun 2005
 
Posted: 2019-10-30 16:49
Have read chapter on electronic trading.

Example of fitting model to past prices which fails to perform on future prices with embedded mechanics of (past again) risk aversion resembles banks a lot.

It is worse - risk crowd imposes harder limits (even cultural) than a single person.

deeds


Total Posts: 459
Joined: Dec 2008
 
Posted: 2019-10-30 16:52

@nikol - for clarification - fitting model to past prices (not returns)?
Previous Thread :: Next Thread 
Page 1 of 2Goto to page: [1], 2 Next