Let's imagine for a second that I am your manager, e.g. a head of the desk. A good thing to have sometimes, keeps you thinking straight. In the situation described, I would be less concerned with that fact that a model stopped working and is losing money, but with the fact that the losses are unmanageable (according to you). My conclusion would be that the problem is in risk management, not in the system design. So, a couple questions: (a) is it the only strategy you are running? If the answer is yes, it's probably worth developing another few to add diversity to your book. If you insist of running a single strategy, you must decrease your leverage. (b) how do you think your strategy mix performs ex-condition? For example, let's say your model buys SPY every time the President of the US has a bowel movement and holds it to the close of the day. What would be your mean, median, 95% worst and the worst daily P&L be ex-condition - meaning, if you just bought SPY in the morning? If you have multiple systems trading at the same time, try to build a history of asset allocations in the portfolio and see it's ex-condition P&L. (c) are you using some sort of a system for strategy allocation? In general, if you have a few years of backtests on multiple systems, you want to build a quantitative allocator that takes into account correlation between strategy performances, recent performances etc. For example, I know guys who like looking at percentile of consecutive losers for each system (vs the back test) and decreasing capital allocation as that percentile increases.
I have the same thing. A system which profits 50% less on the dow compared to the es. very weird. on the other hands, every index is profitable.
Would be surprised if he did! You are joking right ? Part of the problem with trading, nobody sticks together forms organizations. Just a few people sit around and tell others that they are wrong all the time and they cant do it etc etc.
You made a few mistakes: 1. You didn't backtest long enough. 4 years is not enough. 1 market is not enough. Back test for many years across multiple markets. 2. You didn't do an out-of sample test. If you use 2005-2010 to backtest and work out your strategy, you need to then test a different time period - if you system doesn't work in the out-of sample period, it's probably not a profitable system, but just a data-mined fluke. 3. You assumed the strategy would go on working. You had no method for noticing when the strategy had stopped working. Thus, when it happened, you got caught by surprise. The fact is, markets change, systems degrade and stop working, so you have to be prepared for that eventuality, and have a plan to notice it and respond in timely fashion. Another poster (dom993) already gave a good example of how you can use standard deviation and monte carlo analysis to detect when a strategy is busted. 4. You didn't reduce your size after having a larger than expected drawdown. Any drawdown is one of two things - either a bad run in a still-working system, or the start of a system degrading and no longer working. The longer and bigger the drawdown, the more likely it's the latter (strategy failure) rather than the former (bad luck). Thus, the more you lose, the less you should risk per trade - you must follow a reverse martingale approach once your drawdown exceeds the normal level anticipated by your system's prior results. I must say, as a discretionary trader, I see systems traders make these flawed assumptions and mistakes very often. There is something about systems trading that attracts rigid logical thinkers. And rigid logical thinkers (engineer/scientist/quant types) usually tend to overlook the importance of getting assumptions right. They concentrate mostly on deriving accurate conclusions from a given set of assumptions, but pay too little attention to the critical importance of making sure those assumptions were correct in the first place. A bit of lateral and creative thinking, and healthy scepticism about initial axioms, would go a long way to avoiding this mistaken approach. The best solution is to find a creative discretionary trader with plenty of market experience, and ask him to double-check your system for mistaken assumptions, risk blindness, rigid and close-minded design & thinking etc. Unlike engineering, trading requires flexibility and the ability to detect when the 'rules of the game' have shifted. Markets are not a fixed and static system like bridge-building or unbiased roulette wheel statistics. There are some general risk-control maxims you should consider, they should make your trading approach more robust: 1. Murphy's Law - if something can go wrong, eventually it will 2. Real-world results are usually worse than results in testing. Sometimes, they are much worse. 3. Thus, always plan for the worst case, not for the best. 4. A system is only as good as its assumptions. If you haven't analysed the robustness and validity of the assumptions, then you haven't analysed the system. 5. If you haven't considered every possible future outcome, and developed a plan to respond appropriately, then when one of those possibilities you didn't plan for comes to pass, you will not respond appropriately. Good luck.
That's really interesting, I've never felt good about taking big risks, even when they paid off. How can you not feel bad about the potential big drawdowns and even blowup that you are risking?
So, if someone had told you that nothing would work, you wouldn't have tried trading at all? What if someone said the moon is made of blue cheese? I must say, the idea of assuming something to be true, *just because one person said so*, is a bit strange.
No, that's not the lesson. If that was the lesson, you wouldn't have made this post, because you would have no expectation that the post would be made, the internet would still be here, the world would not have exploded, the universe would not have ended etc. The past is very indicative of the future. But flawed understandings of the past are not very good at predicting the future.
Ever notice when volatility is really high lots of "systems" make a decent profit? Ever notice discretionary traders make a decent profit when volatility is high ? Ever notice discretion lose more when volatility is low ? Despite selling volatility- because they sell low Vol. when Vol. is low and they ultimatley get chopped and buy it a little higher, or Vol. plain explodes for a week and they lose. Result, Volatility is usually the one filter that drives this whole business and ET board OR pure trend trader/investor, entire different game 6 months to years outlook.