Profitable, Backtested Trading System? I'm Bored

Well folks, we're back for another round. If you love technical analysis, if you hate technical analysis, or if you just enjoy a good debate about trading, lean in for a good time! If you are unfamiliar with the context, Adam Grimes of SMB Capital, an enormously successful trader, and I have friendly debate going - all in good fun, of course - to increase awareness about trading, investing, and technical analysis in particular. In my

, I challenged Grimes' six-part

, in which he suggested that “every edge we have as technical traders comes from an imbalance of buying and selling pressure.” Many traders who use technical analysis swear by this belief. Because I knew that this belief is central to technical analysis, I specifically countered that traders cannot know if there is "buying and selling pressure" in realtime because of the enormous and inefficiently-reported influence of outside money. (If you are lost, get caught up

.) In response, Grimes

"thought processes behind quantifying the kind of trading edge that may or may not exist around a specific kind of imbalance." Specifically, he argued that thinking about a possible supply/demand imbalance in stocks, creating a trading system based on that idea, and testing that idea on historical data could be a legitimate way to develop a consistently profitable trading strategy. Phew! If that sounds like a lot of in-depth debate so far, well, it is. We are getting at the heart of technical analysis, ladies and gentlemen, and I hope you brought your boldest cup of coffee to your computer for maximum enjoyment. I would like to start by getting a few nagging points out of the way, and then I can move along to the more important critique. Maybe it is just because I am a stickler for good debate, but I hate leaving the cherry pickings on the tree when they are just so tempting. In self-restraint, I will limit myself to two pickings. First (and I hate to mention it, but I must), any method that allows shorting stocks exposes the trader to infinite risk and, therefore, should always be praised with caution. Grimes' method allowed for short selling, so I must hit a sucker punch here. No matter how unlikely, the possibility always exists that a shorted stock could be halted, only to gap up to any level for the reopen of trading. For (an extreme, but possible) example, a trader with a $100k account shorts $100k of ISRG and "risks" only 2% of his account on the trade, entering a stoploss order at a price where his loss on the trade would hit -$2k. Despite his prudence, ISRG is halted for trading, announces a cure for cancer, and the stock is unhalted 500% higher due to unlimited demand for the cure. The trader covers his short position and is impoverished for life. Again, I hate to mention this shortfall of Grimes'

, as I know that he did not really mean to present a complete trading system in the first place; but the cherry was ripe for picking, and I apologize for giving into the temptation. Second, and briefly, any method that does not show maximum and peak-to-valley drawdowns should also be viewed with caution. Sure, Grimes' 64% win ratio and 3.2x greater profit-than-loss do sound great; but without the drawdown figures, it is impossible to determine if his system actually controlled for risk to the trading account. Just two cheap shots, like I promised. Now that I am done cherry picking, I would like to move on to my core critique. Before doing so, however, I would like to acknowledge how much hard work went into the development and testing process of Grimes' response. He sent interns on multi-day projects, crunched some serious numbers, and obviously spent some deliberate time in reflection. I would be remiss to not tip my hat, not to mention offer a sincere extension of gratitude. To Adam Grimes, the readers of his blog, and all at SMB Capital that take time to think through the deeper issues in trading, I salute you! Nevertheless, although a large amounts of work went into the testing process, Grimes' response and trading system are not terribly interesting to me. As the title suggests, I am bored. Below, I argue that Grimes' response should not be very interesting to you, either. An odd spin on an argument: "I'm bored." I admit it. I brought my coffee to the computer today, though, and maybe you should, too… because my reason for boredom might deserve just one extra sip of that brew before you continue. Ah, java. OK, so here we go. Now, there are plenty of trading methods that have extraordinary historical profitability. Let's start there. We have all seen advertisements for trading systems with "90% accuracy and minimal drawdowns" (sometimes ad nauseum for those who have been trading for a few years). No matter how fantastic the claims of these systems, however, their historical performance does not mean that they will actually provide a legitimate edge for profitability in the future. Although backtested results are interesting, they do not prove that the system actually has an edge. Grimes suggests that because his system has excellent historical performance, then it might have an actual non-random edge. I am not convinced, nor am I very interested. Specifically, I am not interested because there are thousands of trading systems with splendid historical returns, just as there are lottery-ticket-picking systems that occasionally win the lottery. Likewise, if I watch enough casino players for long enough - each one following a "system" for playing the slots - I would undoubtedly see a jackpot winner every now and then. Should I, then, approach the jackpot winners and ask them for their "secrets"? Grimes' fallacy is a rather common one and is repeated constantly throughout the world of technical analysis. In a very complicated fashion, he has essentially committed the logical fallacy of

"it does not follow." Because of the importance of this fallacy, I feel as though it would be appropriate to spend some time on this point. Let me summarize my critique, for clarity: selecting one trading system (from an infinite amount of possible trading systems) and claiming that because this system would have been spectacularly profitable in the past we should therefore be more inclined to believe it is based on a profitable edge, is fallacious. It "does not follow." It is simply not a convincing argument, despite how convincing it might appear on the surface. If you have found a historically profitable trading system, great, but I am still bored. The problem with this suggestion (and I acknowledge that Grimes was only meaning to suggest, not prove) is that his suggestion ignores the infinite amount of historically profitable trading systems. That is the problem. To claim that "because I can prove my system is correlated to a possible supply/demand imbalance and made fantastic profits in the past, therefore I am exploiting an actual profitable, non-random edge for continued profit in the future" is a

argument. No matter how wonderful a trading system has performed in the past, one must remember that there are literally an infinite amount of trading systems that could produce profits with minimal drawdowns and robust risk-management in the past. In other words, historically profitable trading systems are, in a sense, "boring." They are boring because there are too many others that are equally (or more) profitable. Sure, it is wonderful that this particular trading system is historically profitable, but, well, so are an infinite number of other trading systems! If you are reading this and having difficulty understanding this critique, perhaps an example will help to illustrate. Say you finished reading Grimes'

about trading In-Play stocks, see the results of his sample trading system, notice that it wins 64% of trades and profits 3.2x more on winners than losers, and decide that you should spend some serious time investigating the reason for these splendid results. Well, although I cannot tell you how to spend your time, I would nevertheless caution you that "because his trading system has splendid historical performance" is not something particularly appealing in the world of backtested trading systems. Pick any system, oh goodness where to even start with the list. Sell JPM $10 puts at will. Sell KEY $50 calls at will. Invest in DEST whenever you want. Sell TRAD $20 calls at will. Pick any "scandalously reliable" trading systems from the endless stream of marketing emails. Or randomly add some technical indicators to a backtester, tweak the parameters until it is profitable, and then modify the numbers in tiny gradients until you see how time-consuming infinity can become. This will show you that there are, as you should already know, an infinite number of possible trading systems that provide even better profits with less risk than Grimes' system. Is it starting to make sense? A backtested, profitable trading system is a dime a dozen. Historical results do not prove - nor by themselves suggest - a profitable, non-random edge. Now that I have clarified this important point, I can finally get to my final response. Many apologies go out for spending so much time on such a simple point (a backtested, profitable trading system is a dime a dozen), but I felt as though it was an important point. My final response to Grimes'

is that he still has not supplied an actual answer as to

traders can find a profitable edge in the future markets based on an imbalance of supply and demand.

, I made the argument that the majority of a stock's true supply and demand is invisible, non-instantaneous, unexpected, enormous, and unpredictable. (If you did not read that critique, I encourage you to read it, along with Grimes' six-part blog series,

.) Sure, Grimes has responded to my critique by showing a historically profitable trading system; but I have argued here that this system is a dime a dozen. Because there are an infinite number of systems that are more historically profitable than Grimes' system, he has not proven or meaningfully suggested that his system is based on a profitable, non-random edge for future use in the markets. Sure it might be

to a

imbalance in supply and demand, but he has not proven this or meaningfully suggested it, other than the undefended claim that large overnight gaps sometimes show imbalances in supply and demand. I would continue to argue,

, that the majority of supply and demand is hidden off-the-chart in invisible, underreported, non-instantaneous, unpredictable, and enormous pools of money In closing, I would like to remind readers that I am not suggesting that profitable trading based on technical analysis is impossible. Throughout this series (

,

, and now part 3), I am merely challenging readers (via Grimes) to explain

they can profitably trade using technical analysis. So far, Grimes has not explained how. Can you?

Disclaimer: Benzinga has no affiliation with Adam Grimes or SMB Capital in any way whatsoever. This article was written to solely increase critical awareness of technical analysis, securities markets, and financial concepts. The views and opinions expressed by Adam Grimes and SMB Capital do not necessarily reflect those of Benzinga. Likewise, the views of this staff writer do not necessarily reflect those of Benzinga. Specifically, Benzinga realizes that technical analysis is a valuable tool for many traders and regularly provides diligent technical updates of the markets by experienced technicians for the benefit of these readers. UPDATE:

that merely one type of off-the-chart trading - dark pools - accounted for 1/3rd of trading in December 2010.