Projection. I think you mean YOU are destroyed psychologically by such things. "I am engineer by training but I consider my poker experience at least as relevant." I have already been through the learning process of emotionally handling the variance associated with small edges. In the poker games I play I win about 35% of the time (50% would be a big step up), but just like the goal in winning trading, my winners are sufficiently bigger than my losers and over 500+ games my edge is very likely to make me a net winner. I've been through the whole mental/emotional process associated with big downswings, I guess you folks call it drawdown. The "Do I even know what I'm doing? Was I just getting lucky til now?" self-doubt, hand-wringing, and anxiety. About the tenth time, you get over it. Your downswing corrects, and confidence takes over. Dealing with variance to me is old hat. Getting an edge with 50% winners would be a small degree of psychological relief, not destruction. In fact, I often read with disbelief a lot of the talk that traders engage in about how tough they have it with drawdown and small edges and psychological discomfort. Do you folks have any idea what blackjack players go through when pressing <1% edges against the casino? Their equity curves would make most traders throw up on their shoes, and some of these guys make a living like that! There was an older thread that caught my interest awhile back called "Systems development with acrary." I highly suggest you go back and read it carefully. What he described there was more or less Gambling 101 - there isn't a concept he talks about there that the educated gambling sharp doesn't already understand and use. Assuming he was credible, he was also a very successful trader over the long term and the thread ended with him being offered "obscene compensation" for a head trader position. To me, this is evidence that my gambling knowledge is not irrelevant to trading as you are asserting. Now, back to my trading journal! I've just launched my second system, and I'll be doing some correlation studies soon to examine how the hybrid system performs as compared to each system independently. The design of the two system happens to be such that they can never trade against one another. BTW, Day 4 turned sour for me on my open trade. My trade history so far: S1 short, -$70 S1 short, +$230 S1 short, +$240 S1 long, -$170 S2 today, I'm long, +$590 and open where S1 is "system 1" and S2 is "system 2". Cheers, Fletch
Agree. I've been an engineer for 26 years now. I also have been running autotrading systems of my own design for several years, and thousands of trades. They have a positive expectation over up and down markets. I'm also up about 25% this year. I know, not much compared to most of you supertraders. But I did it on an average account exposure of less than 50%. As an engineer, I get things done. When I write systems I backtest thoroughly, then I get them out there and trading and I analyze every single trade in Excel (my system writes all the trades to my spreadsheet). Will my stuff always work? Probably not, but I'll adjust. Always have. Sorry for highjacking the thread. But some engineers CAN make money in the markets! Good luck to all.
I also have a master of financial engineering degree, and we do modeling and simulation for financial instruments. Financial engineer is still an engineer, right? By the way, since when do we think engineers are people that over-analyze things? To me, the first desciprtion that pops up on mind is problem-solver. You have a problem, and you throw it to an engineer. He or she may not solve it elegantly, but you can be sure you will get a working solution.
Charlie Dow you are sounding like a Scorned wife. If your such a know it all, what is you edge in the market? tea leafs? Fletch I am curious about your 2 systems. You can give a detailed overview without give away any "propriety" info.
I can give a little overview, sure. Both S1 and S2 are designed to target specific intraday timeframes. S2 catches moves at the open, and S1 catches afternoon trends. I am currently not trading mid-day, but I have a third strategy that I'm developing for that timeframe that looks quite different from S1 and S2, which in some respects are quite similar. The systems were designed by doing some fairly exhaustive searching for statistically significant correlations, based on a whole lot of educated guesses about what might correlate, followed with a lot of trial-and-error to see if any of those guesses had merit. A few seemed worth investigating with a strategy to see if they were exploitable. In the end, the resultings systems are quite simple, so it's hard to go into too much detail without simply giving the systems away. Both systems are in the ballpark of 50% winners, an EV of about $70 a trade on ER2, and profit factors in the 1.5 to 2 range. Both systems trade 125-150 times a year. I'd like to get my trading frequency up from here so that I can start accruing statistically significant results over shorter timeframes. But, I've got what I've got for now, and will have to be patient. Fletch
Sure sounds to me like you know what you are doing Monitor your real time P & L's in excel. Then create a chart to monitor the equity curve. If the drawdown exceeds (percentage wise) what your hypothetical testing has shown then I would start to worry. This works for me. Good luck
Not a bad day on net, but I gave up some early gains with my S1 trade in the afternoon on my first "unforced error." This afternoon somehow I forgot to kick in my automation before my trading window started. I noticed about 5 minutes late, it was a long trade and the signal had just evaporated. I thought "what the heck, 5 minutes can't possibly make a difference" and bought manually anyway. Of course, I picked an almost perfect top (long, of course), and gave up an extra $100 on my trade. In the scheme of things, it's probably insignficant, but it was a lapse of discipline and I don't like those. From my experience at the table, I know that those "I probably shouldn't do this, but..." moments can separate the winners from the losers. Recap: (Day - S1s, strategy 1 short, etc.) 1 - S1s, -$70 2 - S1s, +$230 3 - S1s, +$240 4 - S1l, -$170 5 - S2l, +$480; S1l, -$200 Only about 500 more trades to go before I know where I stand. Shazbatz: Your suggestion is a reasonable one for monitoring P/L and throwing a yellow flag when actual drawdown exceeds max in the backtest. I have a variation of that method that I want to talk a little about soon, that draws on some experience I have quantifying confidence in edges in the presence of uncertainties. Fletch
Tonight I'm doing a few quick calculations on the subject of risk and edge uncertainty. Let's start with risk of ruin. The baseline calculation is very straightforward: for a given strategy, backtest over a reasonable sample of trades, and consider this representative of the "pool" of possible trades the system will generate. We will account for uncertainty in our edge later and the implications for uncertainty in our risk of ruin, but this is the starting point. The risk of ruin calculation from there is very simple: using a Monte Carlo approach, select trades randomly from the sample and transact from a starting bankroll (sorry for my crass gambling terminology) for a large number of trades, say, 2000. We should choose a number big enough that the results don't change much if we continue to make it larger. Since we have an edge according to the trade pool distribution, we expect that if we aren't broke after enough trades we'll be out of danger. So let's start with some sample numbers. I have a mediocre system I'm playing with on ER2 that backtests with an EV of $54 per trade with 52% winners over 350 trades on two years of 5 min data. For a $5k account, this baseline calculation gives a ROR of about 0.2% - 0.3%. I'm not taking into account margin calls here, just a zeroed balance (as if you could put up additional margin but just didn't want to lose more than $5k before calling it quits.) Ok, so let's go trade! We're safe as can be! Well, not so fast. The backtest data may or may not represent the real behavior of the market, of course. Sometimes a system "gets lucky" and shows great returns for a timeperiod that aren't sustainable. In fact, we probably have optimized our system in such a way that it's showing the "luckiest" response we can find over our sample. The next question I'm going to ask is: how likely are we to see the mean result of $54 a trade if our trade samples are coming from a "true" distribution that is worse than our sample would lead us to believe? In other words, how likely is it that our backtest is getting lucky? I don't want to get into the inevitable squabbles about how it is impossible to know the "true distribution" or whether such a thing even exists. Believe me, I have had that debate countless times in the poker community and I'm not going through it again here. We're going to make some assumptions for this exercise. Take it or leave it. So we have to construct some "true" distribution of trades, given only our backtest data which we believe provides a basic outline of the distribution, but may be a "lucky" peek into the real underlying distribution. Let's start simple. I'm going to take the sampled distribution and for every trade I'm simply going start discounting it by a constant amount. I'm just shifting the distribution to the left by a constant amount. You could get fancier, but I won't bother for now. We're then going to start simulating trade series of the same sample size as our backtest, and we're going to see how likely it is that our $54/trade EV was the result of a lucky sample from a less profitable "true" distribution. Clearly if we use the original distribution, we are going to get $54/trade or more exactly half the time in our simulated runs. If our model distribution is worse than that, we are going to get decreasing probability of seeing the $54/trade EV. Here's a sample curve: This is telling us that there's a 15% chance that although our "true" EV is actually $34, we've just gotten a lucky sample over our 325 backtested trades, and we're fooled into thinking our EV is actually $54. So let's relate that back now to the risk of ruin. Since our edge is smaller, we know our risk of ruin is higher. For each of those discounted distributions, we can look at an implied risk of ruin for that distribution by repeating the process described earlier for a sufficiently large number of simulated trades. Here's how that works out: Initially we computed our risk of ruin at 0.25% on our $5k bankroll with this system. As we can see here, that's a very optimistic estimate. It's not terribly unlikely that our risk of ruin could be in the 5-10% range, instead of fractions of a percent, where we'd like it to be. And now here's a key thing. Just eyeballing now, we can consider the product of the two curves in the second chart in terms of a weighted average of ROR based on the likelihood of each assumed "true" distribution. Clearly then the rate at which the first curve "rolls off" is important in protecting us from those disastrous risks of ruin towards the right of the chart. That rolloff rate is a strong function of the sample size. Believe it? Let's try again, only this time we are going to use a backtest with 53 trades in it. 53 trades isn't a lot, but it should be good enough to get an idea, right? Let's try it. Here's the chart with a 53 trade backtest: Now we're a 1 in 4 shot to have a disastrous ~10% risk of ruin! As for me, I pass on that gamble, and now I have a pretty good basis to face facts that a 53 trade backtest doesn't mean much. 325 isn't perfect either, but it's certainly more confidence inspiring than 50. 50 trade backtests belong in the round file as a rule. All of this is just back of the envelope estimates to develop a general feel for what you're dealing with wrt risk of ruin and confidence in an edge from a backtest. Again, these estimates are really best considered as lower bounds, not upper bounds on risk and uncertainty. There are lots of alternate ways to generate trial probability distributions, and you could even generalize this method to account for non-stationary distributions if you really wanted to, but that's as much detail as I'm going to get into for now. Cheers, Fletch
Fletch, I couldn't attach this to a PM or I would have. This is what I was talking about. This chart is a Swing Volume Bar Chart. Everything on the Chart was produced in real-time with the exception of the Buy, Sell, Buy Reverse, Sell Reverse & Exit tags. Those are the outcomes of the green line oscillations on the price portion of the chart. The vertical line on the price portion of the chart are session breaks (day breaks). On the bottom is the momentum indicator that validates the Trend. The Green Line and Trend Tags are generated in real-time as well. This chart depicts a typical month period of time. Would you say that the accuracy would be over 50% "IF" this could be generated in real-time.