Answer to whole thread: Depends on the stability of the correlation.
Making money with a losing strategy
- Thread starter jcl
- Start date
There are two reasons to complicate and mix it up. Automated strategies can achieve maybe ~100% profit in a good year. But the same strategies bundled in a portfolio can make a multiple of that, up to 300% or 400% per year. Profit is here defined as return per required capital.
The second reason is that the profit by a single strategy is highly uncertain. No one really understands why one strategy is profitable and another one is not, or why some strategy suddenly stops generating profits. This is different with a portfolio. Profit maximizing with a strategy mix is mathematically straightforward, the theory is understood, and there are no uncertain factors.
===============
JC;
I think you maybe mostly are on the right track.
Meaning i prefer 3 or more income streams[you may call them systems]. So 3 , maybe 2 done well are better than 1.
Not a prediction, but another reason one may want to include[system b] is -10 % max dd can quite good, long term.
[a]I agree with earlier trader,''not mixing it up''like LTCM.leverege...
[b=system b]Not exact math, but if long GM or BSC, or AMR is losing -10% to S&P /DOW gaining 20% then no.
c]Grain markets are not my main focus, but interesting that grain focused co trader/investor /manager on cover of FUTURES ,June mag used your exact same numbers.80%good trending years;10% profit goal lately...
[F] Of course if one is real young;
coming back from LTCM leverage losses is a distinct possibility.LOL
You happen to be mistaken on two counts
First, the days of selling anything for $10,000 related to trading ended years ago. There aren't any cash-rich newbies willing to toss out $1,000s every which way on a whim. This is 2012, and it's a whole new world in the trading realm.
Second, a losing strat is a losing strat. It cannot smooth your equity curve in a positive way. The peak highs will be lower and the peak troughs deeper. Who cares about what happens in between?
That's not much different than common assumption to reverse-engineer a losing strat to make it a winning strat by fading. Logical deduction, incorrect assumption in reality
I give you a very simple example.Quote from austinp:
Second, a losing strat is a losing strat. It cannot smooth your equity curve in a positive way. The peak highs will be lower and the peak troughs deeper. Who cares about what happens in between?
That's not much different than common assumption to reverse-engineer a losing strat to make it a winning strat by fading. Logical deduction, incorrect assumption in reality
Assume strategy A has 50% win rate, average win $200, average loss $100. Strategy B has also 50% win rate, average win $90, average loss $110. B is a losing strategy. Assume that both strategies do 100 trades per year and are 100% anticorrelated, i.e. the correlation coefficient is -1.
Now a little statistics for calculating the drawdown and the profit. We assume the trades are normal distributed and the probability of N losing trades in a row is thus:
p(N) = (50%)^N.
Let's assume a worst drawdown N at about three times the standard deviation of the Normal curve. 3*sigma corresponds to a probability of 0.0027. Thus,
(50%)^N = 0.0027
=> N = log(0.0027) / log(50%) = ~8.5
Thus, strategy A has an average worst drawdown of
8.5*$100 = $8500
and thus an annual profit of
(50*$200 - 50*$100) / $8500 = 58%.
For simplicity's sake I've omitted margin and commission here, so the $8500 are the required capital for trading the strategy. Now the same for the compound strategy of both A and B. As they are 100% anticorrelated, B always wins when A loses. The average worst drawdown is now
8.5 * ($100 - $90) = $850
and the annual profit of the compound strategy
(50*$200 + 50*$90 - 50*$100 - 50*$110) / $850 = ~470% !!
Adding a losing strategy increases the annual profit from 58% to 470%.
This is an extreme example, as normal strategies are not 100% anticorrelated, but you see that even a slight anticorrelation of a losing strategy can remarkably increase the profit of a portfolio. No higher math and no paradox involved.
Your assumption of 100% non-correlated is flawed... unless you are taking both sides of the same general instrument. Then all you've done is fully hedged a trade, and the term "hedge" truly means "no further profit potential possible" in reality.
There are no markets 100% diverse all the time.
Secondly, a losing strat is a net-loss at year's end. Whether you blend that or run it alone, the same result in part or whole is negative outcome. So there is no way to blend a losing strat with winning strat and experience greater degree of net profit than the winning strat alone.
One can bend statistics anyway desired to curve-fit assumptions. But in the end, systems trading results in real-time is a whole different reality than mathematical equations.
There are no markets 100% diverse all the time.
Secondly, a losing strat is a net-loss at year's end. Whether you blend that or run it alone, the same result in part or whole is negative outcome. So there is no way to blend a losing strat with winning strat and experience greater degree of net profit than the winning strat alone.
One can bend statistics anyway desired to curve-fit assumptions. But in the end, systems trading results in real-time is a whole different reality than mathematical equations.
fwiw: I spent two long, full years of my life in the systems writing world, complete with leasing some 3rd party software for $3,000 annually to run computations until the proverbial cows came home. I often had three computers running different variations of the same concept data-mining all night.
To keep things simple, if you add a net-loss strat to any basket of net-win strats, the end result will be worse in all regards. You may have brief periods of lesser drawdown, but you will also have periods of more extreme drawdown when the winning strats naturally draw while the losing strat piles on the losses.
Losing strats are garbage. Period. All of the time and effort expended to create them and any mental attachment to them is wasted. Delete, discard and move on. A silk Prada cannot be made from Iowa sow's ears
To keep things simple, if you add a net-loss strat to any basket of net-win strats, the end result will be worse in all regards. You may have brief periods of lesser drawdown, but you will also have periods of more extreme drawdown when the winning strats naturally draw while the losing strat piles on the losses.
Losing strats are garbage. Period. All of the time and effort expended to create them and any mental attachment to them is wasted. Delete, discard and move on. A silk Prada cannot be made from Iowa sow's ears
Quote from austinp:
fwiw: I spent two long, full years of my life in the systems writing world, complete with leasing some 3rd party software for $3,000 annually to run computations until the proverbial cows came home. I often had three computers running different variations of the same concept data-mining all night.
To keep things simple, if you add a net-loss strat to any basket of net-win strats, the end result will be worse in all regards. You may have brief periods of lesser drawdown, but you will also have periods of more extreme drawdown when the winning strats naturally draw while the losing strat piles on the losses.
Losing strats are garbage. Period. All of the time and effort expended to create them and any mental attachment to them is wasted. Delete, discard and move on. A silk Prada cannot be made from Iowa sow's ears
I respect your accomplishments but I don't think you understood what jcl was firing at.
This isn't a hedge in the traditional sense of the word, in that it's only aim isn't to reduce maximum losses but to also ensure against a dip in performance in one strategy against an entire portfolio.
I agree in principle that losing strategies are garbage and should be discarded immediately. Time would be better spent looking for winning strategies.
But remember that what you call a losing strategy could be a winning strategy in a different market. And by that I mean, when markets turn on themselves and in so doing, winning strategies become losing ones and vice versa.
jcl's point is that a large collection of uncorrelated to negatively correlated strategies would theoretically ensure a profit regardless of a change in general market direction.
The math's does make sense (apart from the small mishap in his last post), and it would be stupid to disregard it as statistical ramblings.
To add to that point, even though it would be impossible to amass this collection as stated, It could be one's strategy to try and find just enough of these to add to their portfolio to return a healthy profit. After all, It wouldn't be the end of the world if it returned a consistent, unshakable but smaller 10% profit instead of a rather ambitious 470% profit that I saw mentioned in a previous post.
However, on the unshakable note.
To put a question to jcl, what would be the effect on this portfolio in the event of a major black swan event?
My feelings is that as fear grips the market, rules begin to change,
Correlations knock into line. Where you were previously sitting comfortably with vague negatively correlated and uncorrelated instruments trading at a profit you would soon find all your instrument begin trending in the same direction and at that point you would be at the mercy of the market.
My prediction: it would likely trigger a host of margin calls, while a good portion your portfolio descends into a prehistoric bloodbath.
I do sincerely hope that is not the case.
It is indeed so. Ralph Vince has analyzed the correlations of a collection of assets over time, and has found that in situations when prices sharply decline or rise, correlation coefficients rise also. This would probably destroy a portfolio of carefully selected uncorrelated components. You need to observe the correlations for detecting a black swan event, and need to keep reinvestment factors well below the theoretical optimum.
To avoid misunderstandings, probably no one would really add a losing strategy to a portfolio, even when this would generate more overall profit. My point was just that correlations are as important, or even more important than the performance of a strategy. It is not very difficult to find negatively correlated strategies that are not losing.