Will somebody please do this research? / has anyone already?

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.
AFJ Garner
Roundtable Knight
Roundtable Knight
Posts: 2071
Joined: Fri Apr 25, 2003 3:33 pm
Location: London
Contact:

Post by AFJ Garner »

Chuck B wrote: The game has changed from simply testing a system on lots of data, portfolios, position sizing algorithms, etc, etc, and then putting a robust method into production trading. That was the 1990s trend.
Other than back-testing, it is a little difficult to know what else one can do. All that you say is true: can your method ride out the volatility and profit? But other than back-testing, paying particular attention to the more recent past, what can be done?

If one is a believer in absolute determinism, given enough computing (infinite?) power even apparent "chaos" might arguably be predictable. Assuming that determinism is a fact.

It would certainly make the weather forecast more reliable.

But, erm.........I am a little stumped how to put it all into practice.
trackstar
Roundtable Fellow
Roundtable Fellow
Posts: 76
Joined: Sun Mar 09, 2008 11:03 pm
Location: Houston Texas

Post by trackstar »

I would argue that not much has changed. I think its likely that the volatility of volatility has increased. But it could be a function of more participation in general.

I'm not sure I agree much with the statement that Trendfollowers are becoming so common that the edge is wearing down. Central banks are intervening more than ever with larger amounts of money than ever. They don't use trendfollowing tactics, quite the opposite really. Take a look at the Swiss and Yen markets recently. Talk about fading rallys......

Those with the best money management strategies will continue to win as they always have.
BuyHigh SellLow
Roundtable Fellow
Roundtable Fellow
Posts: 50
Joined: Wed Apr 27, 2011 12:46 pm
Location: U.S.

Post by BuyHigh SellLow »

adamant wrote:In an interview in 2000, Jim Simons stated that commodity markets used to trend nicely but don't any more. He was wrong (see gold for a good example). The day human psychology changes and people stop trading based on their emotions, we will have a problem. I don't see that happening any time soon.

Do markets become more efficient over time? Probably. Will there still exist ways to exploit the inefficiencies created by human emotion in 20 years? Probably. Is trend following my best bet with regard to consistently making money in the market? I am fairly certain of that.
That interview sounds familiar. Was that the same one where Simons said, "trend following has been mined dry."

Until I see a given TF strategy setting new, horrendous draw down records in a new way, I am not concerned. A TF friend of mine (smartest guy I know) said it best:

"If you don't do TF, what then? Ignore trends? Sell new highs and buy new lows? You have no choice."
Chuck B
Roundtable Knight
Roundtable Knight
Posts: 481
Joined: Thu Apr 17, 2003 6:34 am

Post by Chuck B »

Some central banks have gotten MUCH smarter than in years past however. They know they need to make price action reach certain levels to trigger exits in the big guys for example. How many long term trendfollowers are going to remain in the position when a 12ATR move from the prior close against them is in progress? 14ATR? 250 day opposite extreme?

If I was a central bank desk intervening, I'd have a plan worked out to cause maximum pain to the large vested speculative money, and when after I've intervened, slowly, methodically enough to have the market approach some of those key levels, 20 day low, 50 day low, 3 to 5ATR off prior close, etc, that's when I'd light the fire on my intervention. Drop the bunker buster bombs x 50. My goal would be to expose all those interests to the largest slippage they've ever seen on their exit orders, and if I get real lucky my price action will trip them onto a "trend" in the opposite direction about the time I let up on the intervention. :wink: Heck, I'd be the Stewie of central bank desk managers. :D :D
rabidric
Roundtable Knight
Roundtable Knight
Posts: 243
Joined: Mon Jan 09, 2006 7:45 am

Post by rabidric »

@Chuck B

All your posts in this thread resonate very strongly with my own thinking. I too read widely in the various fields of General Systems Theory, Complex Adaptive Systems, Evolutionary Dynamics, Heuristics, some econo-physics, theories of Warfare etc.

Your idea of capacitance/inductance of pits, is a new way of expressing "it" to me. I always thought of "it" as "viscocity" of human trading(auction) interaction, but that probably doesn't quite nail it as well as your words.
Thanks.

P.S. I have dreamt about what i could do with a central banks power, but it wouldn't be just mere intervention. Think about what kind of economic warfare and profit you could make against another unsuspecting country which wasn't prepared to combat a offensive profiteering Central Bank , helmed by a balls out trader? 8)
LeviF
Roundtable Knight
Roundtable Knight
Posts: 1436
Joined: Mon Dec 22, 2003 12:24 pm
Location: Des Moines, IA
Contact:

Post by LeviF »

BuyHigh SellLow wrote:"If you don't do TF, what then? Ignore trends? Sell new highs and buy new lows? You have no choice."
My thoughts exactly.
Chuck B
Roundtable Knight
Roundtable Knight
Posts: 481
Joined: Thu Apr 17, 2003 6:34 am

Post by Chuck B »

LeviF wrote:
BuyHigh SellLow wrote:"If you don't do TF, what then? Ignore trends? Sell new highs and buy new lows? You have no choice."
My thoughts exactly.
Just realize that the above statement is simply a belief and not the end-all truth. Recognizing it as just a belief, perhaps a belief that has been profoundly useful in the past (and hence the adoption of the belief), but in the end just a belief. It may be yet again profoundly useful going forward too. Who knows? Nobody for sure.

However, if that belief defines your total internal map of the markets, then recognize that, this one, like most other beliefs are all limiting in that they can affect your ability to see a different market map (or even believe there could ever be a different market map).

If you pay homage and respect to that belief, but at the same time adopt another belief: "if I feel there is no other choice, many billions of dollars worth of other trendfollowers likely feel there is no choice either." Now see what new market maps your brain can hypothesize given that new belief. It might just open you up to a whole new market map you never dreamed you could think up. All breakthroughs I've ever made in doing system work have started out this way including the initial shattering of my belief that TF wasn't worthy of my time about 20 years ago, lol. :) I went from that limiting belief to creating a TF system and running a portfolio of markets with it within about 2 years. :)
Chris67
Roundtable Knight
Roundtable Knight
Posts: 1052
Joined: Tue Dec 16, 2003 2:12 pm
Location: London

Post by Chris67 »

if you wanted to break down a study of correlation changes (forget whether its input or output for now) then anyone recommend the best way to do it ?
Say, as a simple example, you have 2 market instruments - the Euro and the Swiss
Say you look at correl over the 1980-1985 period then walk it forward looking at 5 year periods and plot this - does that make sense
The other method I guess - and maybe a stupid idea - but if you have 50 markets and you want to see their correlation changes over time - wouldnt an obvious way opf doingt it simply be to take the correl matrix and add up all the correls for a 5 year period then produce teh same matrix for teh next 5 year period, add up all the data points and if one is higher than the other - overall market correlations are higher ?
Just one concept - I have no problem doing the legwork (it makes a nice break from being chopped in the markets ) if someone can add some weight to these ideas ?
I have the resource of someone cruncing numbers for me at present so happy to utilise and share findings
Chuck B
Roundtable Knight
Roundtable Knight
Posts: 481
Joined: Thu Apr 17, 2003 6:34 am

Post by Chuck B »

rabidric wrote:@Chuck B

All your posts in this thread resonate very strongly with my own thinking. I too read widely in the various fields of General Systems Theory, Complex Adaptive Systems, Evolutionary Dynamics, Heuristics, some econo-physics, theories of Warfare etc.

Your idea of capacitance/inductance of pits, is a new way of expressing "it" to me. I always thought of "it" as "viscocity" of human trading(auction) interaction, but that probably doesn't quite nail it as well as your words.
Thanks.

P.S. I have dreamt about what i could do with a central banks power, but it wouldn't be just mere intervention. Think about what kind of economic warfare and profit you could make against another unsuspecting country which wasn't prepared to combat a offensive profiteering Central Bank , helmed by a balls out trader? 8)
The first country that came to mind when I read your last paragraph was China. The second thought was if global relations (of whatever form between whoever) breakdown at some point, I would assume that the front end part of that conflict would involve economic warfare of some sort. Perhaps a combination of attacks of "secure" network data/systems coupled with a currency and/or sovereign debt market action. Ugh...
Moto moto
Roundtable Knight
Roundtable Knight
Posts: 427
Joined: Mon Jun 01, 2009 4:12 am
Location: once again in the UK

Post by Moto moto »

obvious question.....if trend following is dead.....what is alive?

It appears to me that a lot of methods are failing at present.
AFJ Garner
Roundtable Knight
Roundtable Knight
Posts: 2071
Joined: Fri Apr 25, 2003 3:33 pm
Location: London
Contact:

Post by AFJ Garner »

With the greatest respect and with all due allowance for my own shallow and unhelpful response to Chris, this thread is actually going nowhere.
On the one side, various parties have said "trends will remain for ever" on the other people have rightly said that volatility may make those trends difficult to follow.
Much has been stated about changing markets and technological advances in dealing (with mention of some uncomfortable consequences such as wicks).

No suggestions have been made by anyone as regards to what action to take. If traditional back testing is indeed dead, then what is the new paradigm?

If the situation is as glum as some suggest, then remove what you have left from your fraudulent broker accounts and stuff it under the mattress.

This is typical of the glum talk at the bottom of one of these ghastly market melt downs. Who knows, perhaps trend following is indeed dead this time round. But like Lazarus or the phoenix, it may yet surprise us. And if it doesn't what then do people recommend?

Concrete ideas are always so much more uplifting than cries in the dark.
Chuck B
Roundtable Knight
Roundtable Knight
Posts: 481
Joined: Thu Apr 17, 2003 6:34 am

Post by Chuck B »

Moto moto wrote:It appears to me that a lot of methods are failing at present.
One that failed for me was believing my Seg-US cash in two MFG accounts was actually secure. Hard to trade when your FCM is stealing from you and your accounts are frozen.
AFJ Garner
Roundtable Knight
Roundtable Knight
Posts: 2071
Joined: Fri Apr 25, 2003 3:33 pm
Location: London
Contact:

Post by AFJ Garner »

Chuck B wrote: One that failed for me was believing my Seg-US cash in two MFG accounts was actually secure. Hard to trade when your FCM is stealing from you and your accounts are frozen.
Seconded.
Chuck B
Roundtable Knight
Roundtable Knight
Posts: 481
Joined: Thu Apr 17, 2003 6:34 am

Post by Chuck B »

AFJ Garner wrote:With the greatest respect and with all due allowance for my own shallow and unhelpful response to Chris, this thread is actually going nowhere.
On the one side, various parties have said "trends will remain for ever" on the other people have rightly said that volatility may make those trends difficult to follow.
Much has been stated about changing markets and technological advances in dealing (with mention of some uncomfortable consequences such as wicks).

No suggestions have been made by anyone as regards to what action to take. If traditional back testing is indeed dead, then what is the new paradigm?

If the situation is as glum as some suggest, then remove what you have left from your fraudulent broker accounts and stuff it under the mattress.

This is typical of the glum talk at the bottom of one of these ghastly market melt downs. Who knows, perhaps trend following is indeed dead this time round. But like Lazarus or the phoenix, it may yet surprise us. And if it doesn't what then do people recommend?

Concrete ideas are always so much more uplifting than cries in the dark.
Reread my post up above about beliefs. I hope you're not dumping me into some basket saying long term TF is dead since that's not the belief I have. Instead I try to have a set of useful strong beliefs about markets that are lightly held. :) Hence all my comments about assuming things about the future and preparing for them, how you'll handle them, whether you'll ignore them, etc. I'd rather be prepared and have a rehearsed plan rather than blindly believing I've found "the answer" forever more...
AFJ Garner
Roundtable Knight
Roundtable Knight
Posts: 2071
Joined: Fri Apr 25, 2003 3:33 pm
Location: London
Contact:

Post by AFJ Garner »

Chuck B wrote: I hope you're not dumping me into some basket...........
No Chuck I would certainly not do that and I appreciate your views. These are very difficult times indeed and a difficult trading year has been enormously exacerbated by the appalling and ineptly handled collapse at MFG.

At times like these I find it all too easy to give into bleak pessimism and I think many of us are feeling a little that way at the moment. It can be every bit as damaging as blind optimism.

My only real point is that one's beliefs need to be translated into a plan of how to trade for the future. Or not trade as the case may be. I for one do not believe that back testing is dead. You still need to look at past data, apply to it a system, a portfolio and money management routines.

I was merely trying to elicit specific ideas on how such techniques might be supplemented and improved.

The only thing I want to throw into any basket is friend Corzine and assorted ego inflated maniacs who yet again have brought the financial system to its knees after having ripped off enormous salaries, bonii and IPOs,
Chuck B
Roundtable Knight
Roundtable Knight
Posts: 481
Joined: Thu Apr 17, 2003 6:34 am

Post by Chuck B »

I agree, and I don't have any specific answers about testing per se other than to try to look outside the box past data puts us all in and be really creative about dreaming up potential future paths. The more future potential paths I can accumulate, the more market maps I can dream up, the more resourceful I hope my efforts can be. In the end, maybe the answer is just to stick with what I've done over the years or some minor derivation of it, but at least I will have expanded my market map hugely through creative research outside the constraints of past data and current system design software.

It's been a while since I've done creative brainstorming sessions, but this whole MFG fiasco has given me a nice break that instead of drubbing around all week I've been working on many new things and actually seeing them in a different light (all of which is probably why I've been somewhat prolific in my posts here on the board this week while being mostly absent from it for years).

Cheers,
Chuck
liaojing
Contributor
Contributor
Posts: 3
Joined: Tue Mar 08, 2011 3:23 pm

Post by liaojing »

I think some of these 6 observations can be solved somewhat easier, like trends reverse quicker. We can adjust our system to adapt to new trend patterns. Just give trades more room at the beginning of the trend, and after accumulate some profit, then gradually tightening the trailing stop, after make some very big profit, simply profit taking, or use very tight trailing stop like previous day's low to exit. Such strategies work pretty good in this year's environment, stay in this year's big trends like silver, gold and bond to the end of trend and give back very little when the trend reverse due to very tight trailing stop.
Some other observations are more difficult to handle, like market becomes more correlated. May need to hold less markets when they all become correlated together. This can reduce the drawdown, but will inevitably reduce profit at the same time. Seems no easy solution.
Some other observation, like market trend less, may have some truth in it. In my opinion, though trend following advocates the idea that TF should ignore the fundamental, I think that doesn't mean fundamental is not important to TF, that only means no one can consistently accurately predict the fundamental, so trend following adopt a passive strategy to simply follow. What TF means is passive follow may be better than subjective prediction. But for trend following to really shine, there must be some Big and Sustainable Fundamental trends behind it. Without sustainable fundamental trends, the price can rise/fall dramatically in short time, but will reverse its previous move quickly due to no sustainable fundamental reason supporting it. Only if the fundamental change is big and sustainable, the trend can stay there for relatively long time, giving TF system to gradually tighten up trailing stop and reap profit. So the real reason TF system can make money actually depends on whether the world have some big and Sustainable Fundamental change. Sustainable is the key and by saying "sustainable" I mean at least stay there for 1 or 2 months. Like the last bonds move, rise dramatically in short time and stay at that level for a few months. So as long as the world is moving, there are some Sustainable fundamental change, TF is sure to make money though we may need to adjust our strategies to adapt to newest trend patterns change. And there will be favorable market conditions like 2008, and there will be conditions where not many sustainable fundamental changes like this year. Just like in some trading book, be relax when the market is favorable and have lots trends; be confident when market is difficult; and be patient and reduce leverage when necessary in challenging environment. As long as the world move/change and there are sustainable fundamental changes, the concept of TF will work forever. Some out-dated TF systems, like original turtle, may stop working, but the concept of TF will never die.
I am pretty new to this forum and just my understanding.
SimJimons
Senior Member
Senior Member
Posts: 31
Joined: Mon Jun 20, 2011 9:59 am

Post by SimJimons »

Isn't the very fact that we're discussing the death of trendfollowing what makes trendfollowing work? I have no suggestions what to do going forward except doing what we have been doing so far. Why waste time speculating about something we will never find an answer to (until it's too late, that is)? Instead, why not focus on the data we have at hand and do the best we can with that, in its traditional sense?

Obviously, I also subscribe to the notion that low correlation is a good thing when trading systematically. Having said that, applying the same system to two highly correlated futures may result in surprisingly different return streams. In addition, the real killer trades usually surface when correlations go to one, as was the case in 2008. So, perhaps the increased correlation lately is only a sign that huge profits are waiting round the corner :P Let's imagine that October 2011 is in fact April 2008 :wink:

Chris67, what do you mean with "Winton's new permanent return profile"? Have they announced that they will be doing something differently, or are you referring to the fact that they have lowered risk to 10%? Also, sorry for posting such a boring answer to your question :?
Last edited by SimJimons on Fri Nov 11, 2011 8:12 am, edited 1 time in total.
Bravochico
Full Member
Full Member
Posts: 24
Joined: Wed Sep 14, 2011 11:46 pm
Location: Southern California, Ca

Follow the money

Post by Bravochico »

Follow the money, ie who has been doing well, and that should glean some insight.

As far as I know, hft traders are doing best over last 5 years.

I'll just list some blink statements.

1. News data feeds have compressed trends
2. News data feeds now have tags on them which can be automated into strategies. Ie, Reuters releases "Italian rates soar". Any of these words can be tagged to execute sell,buy orders in milliseconds.
3. Most hft shops I know are at or near equity highs.

4.The large TFs seem to canabilize each other to larger degree. Fewer large TFs have a larger share of capital in futures.

5.Having a portfolio of 100 futures contracts definitely doesn't provide the benefit it did 5 to 10 years ago.
Chris67
Roundtable Knight
Roundtable Knight
Posts: 1052
Joined: Tue Dec 16, 2003 2:12 pm
Location: London

Post by Chris67 »

unfortunately most HFT shops are not far away from some rather large problems and issues and this type of trading has no longer term track record - bit like nasdaq day trading was really profitable 1995-2000 then they all disappeared - robust ? I think not
Post Reply