Let me try to summarize his position: "I am well aware that some markets do show trends (in retrospect), and allow back tested systems to work. But merely because it worked back tested, why should the markets be so kind as to allow those who can draw a straight line between two points to make money in the future. Mr. Bacon would argue against it in the field of horse racing. Presumably the wisdom of the market is at least equal to the deceptive practices of the trainers and horsemen." He doesnt deny that there are studies that capture trends(like momentum studies), or the Covel system. He argues that -One can't be very confident those results didnt come from randomness -Even if they didn't, one can't be confident the market wont close out that easy money opportunity and the system will stop working Well, I cant argue with that, the first one is a mathematical reality and the 2nd one is also true, systems eventually stop working, system creators will agree with that. For instance I'm quite skeptical of the 'carry trade' inneficiency, I can't see no reason why higher yielding currencies should be able to outperform lower yielding ones forever, the studies show that has been the case. However that could change There is the 'human nature' argument, that says those inneficiencies reflect innate human flaws and thats why the inneficiency wont ever close. However the other side of that is that, the flaw doesnt have to be corrected for most people, only in certain amount of very rich individuals. For instance it only takes one very large hedge fund run by someone with less flaws(say John Paulson) to offset thousands of flawled retail traders. That is because the buying power of the non-flawed investor exceeds the one from an individual flawled retail investor by many times
With regards his criticism of big TF funds, he seems to argue that when those funds get big money their systems stop working and they lose(here the TFers tendency to use small samples works against them as John Hendry lost when he got big, so did a few others). The smart money takes the other side their trades and captures that lost money. I'm not in a position to comment on this however, as I have no data that ever since TF funds started the net $ return to the public has been negative already ajdusting for suvivorship bias and other problems. I'd be interested in seeing something of this kind
I am not so sure the sum of my two books is accurately reflected here, but for the sake of argument let's assume your view of my work is correct. So how does one explain away "The Winton Papers"? That would be David Harding's effort that you can either find in whole on the net or in various PDF files. Try searching for PDF only results with terms like Harding, Winton, Sharpe, etc.
His arguments are overly broad in that they can by used against ANY potential trading approach. Global macro, value investing, merger arbitrage, intraday scalping etc. I see NOTHING in Niederhoffer's arguments that casts any doubt on Trendfollowing specifically. If anything, his arguments give weight to the Efficient Market Hypothesis. Any trader should feel insulted.
I believe you are referring to the paper where he says the Sharpe ratio misses cases where people will make small amounts for a long-time then blow up, while others are swinging a lot but make money over the long-run. Thats a good point but I'm talking about you addressing the statistical tests brought by VN. I addressed the correlation in the stock market(correlation doesnt tell you about the mathematical expectation) I dont seem to recall you making that case in your book but you had no problem saying 'VN blew up' Thats how far I can go however as I have I no background on statistics and have no idea what 'correlograms', 'arima estimates', 'variance ratio tests' are Ultimely TF comes down to a historical inneficiency that some systems are able to capture(like the big TF funds and the momentum studies) but that could be random and not likely to persist or that they could be arbed away but also that could continue to exist. The big TFs are betting that it will while VN bets it that it wont(at least for big sums of money), bottom line is neither party is the position to affirm that science is on their side because the uncertainty is too high
This is true, one cant always rule out randomness which is why trading at the end of the day is about taking a RISK, a risk that you got fooled by randomness and you are trading something that will lose. You look at the data and take calculated risk you believe are likely to payoff then go with it I have a problem with the Covel position of suggesting he must be right because 'this list of names made a lot of money trading a very similar system', its not science its a speculation. He has no background on statistics(that I'm aware) yet tries to make claims like that At the end of the day, trading will not be science but alchemy, success doesnt prove that you will have more success because success affects the future. Making money using a system for 10 years doesnt 'prove' scientifically that the system will work in the future because social sciences are different from natural science, the mere fact that something works draw others to it and it affects that same future How rising awareness that momentum works affects the momentum systems is something I've been thinking about but have not reached a conclusion
I don't think one can ever scientifically prove a trading methodology will make money forever into infinity because we're always "just" looking at past data. In pratical terms there is a simple but elegant solution to such problems (IMO): Reduce the allocation to a respective system by X% if drawdown is bigger than Y%. Re-allocate once it recovers from drawdown. That helps (IMO) living with the uncertainty of a market inefficiency disappearing over time.
With all due respect at some point the "trend followers still around are all lucky survivors" argument is head in the sand. Are we to the point that some people literally explain away Harding, Man, AHL, Aspect, Hite, Millburn, Dunn, Henry, Eckhardt, Turtles, Sunrise, Parker, Rabar, Abraham, Mulvaney, Campbell, etc. etc. -- by simply walking away from their performance satisfied that an explanation of "luck" absolves them of further investigation as to why they make $? That's bizarre to me. But let's face it EVERY college classroom is built upon the foundation that trend followers can't exist. To accept trend following puts egg on the faces of many, many tenured minds around the world.
MAN AHL (packaged in Athena Guaranteed) 20 years of returns. 14% annualized ROR% (net after fees, closer to 20% excluding incentive fees). Max DD 21%. These are absolutely mindblasting risk adjusted returns. Hard to call this luck or a statistical artifact, especially since VN started attacking trendfollowing as a valid strategy over 5 years ago and the results have been nothing but stellar since then.
Find "The Winton Papers". So the 'tests' rule out all the names and performance I mention above as just lucky monkeys banging on the keyboard? Instead of asking me to invalidate these tests, why don't you do a better job of invalidating trend following!