For the sake of the CFA exam, I can “pretend” that markets are efficient, at least in the weak form, and give them what they want to hear. However, in practice, I have a hard time accepting it. For example, how does the EMH explain the 1982 - 2000 bull market? If I tried to generate a stock chart using a random walk generator, I would not see as many “trends” as observed in the market. I’ve been trading the markets with a trend following system since the beginning of the year and have been much ahead of the market (I just follow the trend on major index ETFs like SPY, DIA, QQQQ, etc.) Without going into details, I’ve used a modified version of a trend following system circulating around with well defined conditions when to enter/exit, plus backtesting the results over the last several decades or so (including catastrophies such as 1987, 2002, 2008, etc.) Simply by following the trend, the outperforms the market consistently on a medium to long run basis during bull markets although not by very much… and during bear markets, it makes money and/or keeps losses minimal. Although I only backtested through the end of last year when I first made the system, it worked fine and picked up the end of the rally in late April, re-entered in early September, etc. I’m up over 10% just by following the trend on the S&P500, Nasdaq, and Dow (vs. the indices up in the low single digits right now.) The only time trend following is unprofitable is when you’re unlucky enough to buy/short a security right when the trend is ending… or it incurs a sudden shift to higher volatility right before you enter. And even then, you are given plenty of warning signs before things get real ugly, at least when trading major indices or commodities. I’m interested to hear why according to the Efficient Market Hypothesis, trend following does NOT work. Unless stock exchanges around the world agree to ban disclosing any price data for the recent past, one will always be able to find and follow trends.
EMH is clearly contradicted by events like the October 1987 crash: value (earnings expectations) did not drop 25% in one day. You’ll find none who defend emh on this web site. > I’m interested to hear why according to the Efficient Market Hypothesis, trend following does NOT work. Why do you care? You’ve demonstrated that your system works. Asking us how many angels can dance on the head of a pin won’t impact your success. Get in there and trade! The real answer is: EMH is agnostic wrt the effectiveness of trend trading, so your question’s implicit assumption is off. (They happily coincide if you can demonstrate that fundamental values can trend, which they frequently do. See e.g. http://en.wikipedia.org/wiki/Business_cycle)
Actually, people who publish BS against TA are kinda of desperate scholars in need of research papers. They take really simple data sets and technical rules, to prove that TA doesn’t work. Moreover, how do you determine by backtesting that trend following doesn’t work? By taking the returns between start and end of trend ? No one has proved that trend following doesn’t work, because noone is yet confirmed about the signals of trend start, at least in academia. With least of practical knowledge, they take any arbitrary criteria, like… if more then 10% return in 20 days gives a higher return after that. You can’t classify trends like that, and that’s where real beauty of behavioral analysis lies, that’s the depth of practical experience which let someone decide a trend early, you can’t make rules to enter or exit a trend, it’s about the experience and intuition, research scholars have none of the two! And if TA isn’t working at all, how the hell Renaissance Technologies and D.E. Shaw and many others like them are posting consistently steller returns. I am sure EMH research community would not hesitate to turn that as an exception or outlier or sequence of many random events which someone worked with probability of 1 in a billion, good enough for a research paper. I really don’t have any answers to give to EMH supporters, except middle finger… Go to hell, Outliers beat the market, and that’s not chance!
I have a very good friend who is currently working on his PHD in finance. He and I will often go and grab a beer somewhere where we can be nerds and discuss the market. We often end up in a friendly argument over the effeciency of the markets. His comment is “you can not statistically prove (under a given set of assumptions) why these strategies work.” This is why you do not see research reports from professors recognizing these strategies. There is a clearly defined difference between the academic world and the professional world in high finance. The academic world provides a basic theorectical understanding of the markets, but in practice it is much more dynamic. Keep this in mind when you find a strategy that works for you. Professors follow a given set of assumptions (the most dangerous being that people are rational) and often cannot recognize these situations. Keep trading till this does not work and then find your next strategy!
If I remember rightly EMH theory and random walk are based on daily price movements. The assertion being that prices can wiggle this way or that based on reactions to random information. They don’t look at longer term movements over the course of weeks etc. It could be that trends develop based on affirming information that causes the movement to develop over time in a non-linear way. That way it would not necessarily be mutually exclusive. Otherwise EMH is more academic crap (cf CAPM, VaR, positive economics and anything else that needs a range of unrealistic underlying assumptions to make it work).
Is this a serious question? Trends are always HISTORIC
And? So is the EMH/random walk data. What is your point?
Longcat, congratulations on your positive results for the last 9 months. It has been a great time for trendfollowing indeed. Hopefully, you don’t think that making money in the markets is easy because once upon a time that attitude really hurt my trading. It took me several years to appreciate the difficulty of trading and learn humility. Good luck!
Longcat1982 Wrote: ------------------------------------------------------- > […] The EMH in any strong form is ridiculous. It is indefensible from any number of approaches (practical and theoretical). For example, there was no volatility surface in options before the crash of '87. Implied volatility was flat (the same) across different strike prices for a given expiration. Nowadays, wing options tend to exhibit higher implied volatilities. (This phenomenon is called volatility skew or “volatility smile”.) Which is right? Flat vol or vol skew? Either way, the market is either wrong now or was wrong before. I wish the EMH was phrased more as a limit as t -> \infinity … because the market does get more and more efficient over time. Nonetheless, it never was, is not, and probably never will be 100% efficient.
I don’t think that even the CFAI really expects you as a candidate to “buy into EMH”. Point in case: Emphasis on valuation at level 2 and active management at level 3. Now CFP on the other hand is a totally different story (I blush to even mention the latter designation int he same post as I firmly believe it is nothing more than a glamorized insurance salesman cert). edit Post to add the following caveat: This all really has very little to do with trend following or TA, which in my own opinion adds very little alpha, and there is certainly no dearth of empirical data that supports this opinion. For your academic friends suggest the following paper first published in 2008: Case Closed by Robert A. Haugen and Nardin L. Baker. I wish a could simply post my copy of this here but at the least here is an excerpt of the abstract. I think most of you should find this a compelling read: This article provides conclusive evidence that the U.S. stock market is highly inefficient. Our results, spanning a 45 year period, indicate dramatic, consistent, and negative payoffs to measures of risk, positive payoffs to measures of current profitability, positive payoffs to measures of cheapness, positive payoffs to momentum in stock return, and negative payoffs to recent stock performance. Our comprehensive expected return factor model successfully predicts future return, out of sample, in each of the forty-five years covered by our study save one. Stunningly, the ten percent of stocks with highest expected return, in aggregate, are low risk and highly profitable, with positive trends in profitability. They are cheap relative to current earnings, cash flow, sales, and dividends. They have relatively large market capitalization and positive price momentum over the previous year. The ten percent with lowest expected return (decile 1) have exactly the opposite profile, and we find a smooth transition in the profiles as we go from 1 through 10. We split the whole 45-year time period into five sub-periods, and find that the relative profiles hold over all periods. Undeniably, the highest expected return stocks are, collectively, highly attractive; the lowest expected return stocks are very scary – results fatal to the efficient market hypothesis. While this evidence is consistent with risk loving in the cross-section, we also present strong evidence consistent with risk aversion in the market aggregate’s longitudinal behavior. These behaviors cannot simultaneously exist in an efficient market.
maratikus Wrote: ------------------------------------------------------- It took me several years to > appreciate the difficulty of trading and learn > humility. Good luck! Heh, thanks… 2008 taught me humility as I’m still a bit down from my 2008 highs… in fact, getting owned by the market in 2008 was what prompted me to spend so much time researching and devising such a system last year (and who knew that a buy and hold strategy actually would have worked fine without too much effort.)
> And if TA isn’t working at all, how the hell > Renaissance Technologies and D.E. Shaw and many > others like them are posting consistently steller > returns. I am sure EMH research community would > not hesitate to turn that as an exception or > outlier or sequence of many random events which > someone worked with probability of 1 in a billion, > good enough for a research paper. I really don’t > have any answers to give to EMH supporters, except > middle finger… Go to hell, Outliers beat the > market, and that’s not chance! EMH doesn’t say that you cannot beat the market. In the universe of fund managers, the way returns are distributed there will be definitely some managers who outperform the market over time. EMH just says that the outperformers aren’t winning due to their skill. Regarding DE Shaw/Renaissance, how much leverage do they use? I don’t think a broad market index is the appropriate benchmark for a levered up shop. My issue with the EMH is that, considering that market efficiency is created by the huge amount of active investors, how do we know that the people making up the market are valuing it “correctly”? That is, are people in the market all or even mostly rational?
Just completed the Myth of the Rational market. Excellent read.
a) Market efficiency doesn’t claim that we don’t have trends. We all expect stocks to have long-term up trends, 10-year notes to outperform rolling 3-month bills thus giving trends, etc… The OP claims “If I tried to generate a stock chart using a random walk generator, I would not see as many “trends” as observed in the market”. Did you try a geometric walk with drift? EMH surely doesn’t claim that markets are simple random walks. The difference that you will see is that markets aren’t stationary but EMH doesn’t say they are. b) EMH doesn’t claim that you can’t use information known now to say that markets were inefficient at some time in the past. I can’t pretend to completely explain vol surfaces but they are certainly much richer now than they were in 1987. One natural explanation for that is that people learned that dynamic hedging of equity options didn’t work at all in the crash of 87 and a bajillion times after that. In any event, vol surfaces are “wrong” now and they go through all sorts of regime changes. Implied vol is certainly a highly biased estimator of realized vol. Of course, that’s because all the assumptions in the implied vol models are wrong. Markets being “wrong” doesn’t contradict EMH. I could have made lots of money in 1987 by playing lack of smile knowing what I know now. c) Hedge fund returns don’t contradict EMH because EMH says nothing about all the kinds of stuff that hedge funds can do. I was the risk manager of some unnamed now defunct hedge fund and won a risk management award based largely on manager and model marks. Think that contradicted EMH? Even when they invest in highly liquid markets hedge funds can have fewer restrictions than nearly everyone else (like they can invest billions of dollars quickly, have a large say in corporate management, etc.). Think you can influence markets if, say, you are on one side or the other of 80% of the open interest in all futures contracts traded on some commodity (been there with the full knowledge and consent of all governmental agencies and exchanges)? EMH doesn’t say that you can’t outperform the market by, say, doing a stellar job of running your company. If it did, Steve Jobs would directly contradict EMH. d) Academics are smart people who write papers to get them published for all the academic benefits of getting them published. Almost nobody in academia really cares that much whats in the paper. If a paper is particularly outrageous, more people respond to it and cite it and you get more brownie points. That paper mentioned above is just the “markets were wrong therefore they weren’t efficient” nonsense. Stuff like “The academic world provides a basic theorectical understanding of the markets, but in practice it is much more dynamic” is nonsense. Many academics have deep understandings of capital markets and revolve in and out of finance. Go up to someone like Emmanuel Derman and say something like that (he’s a nice guy and would politely ignore you). e) Beating the stock market for a year or two or ten proves nothing. f) Trend-following absolutely works on big enough size, but it doesn’t work on one market. It’s not too hard to use trend-following to construct “convergence options” or to take advantage of leveraging numerous long-term drifts and diversifying away risks. What you end up beating is something like a compilation of investable long-term drifts and that’s not given by some standard market index.
Nice post JDV.
Great post JDV. Thanks. Anyway, for the past few days, I’ve really enjoyed reading JDV’s posts, and it seems like everyone on AF knows him and misses him while he was gone. That said, as a pretty new member on AF, who’s JDV? (I know it probably doesn’t matter, but I’m just curious.) I searched him and there are more than 7500 posts from him, much more than Numi.
I think the problem most people have with the EMH is two things - they mistakenly assume market efficiency means market stability, and their second problem is they ask-why didn’t the EMH predict the market crashing in 08? Notwithstanding that EMH says precisely that you cannot predict when the market will crash. I don’t know if I buy into the EMH, but it’s pretty hard to refute given its assumptions.
dhyun3 Wrote: ------------------------------------------------------- > Great post JDV. Thanks. > > Anyway, for the past few days, I’ve really enjoyed > reading JDV’s posts, and it seems like everyone on > AF knows him and misses him while he was gone. > That said, as a pretty new member on AF, who’s > JDV? (I know it probably doesn’t matter, but I’m > just curious.) > > I searched him and there are more than 7500 posts > from him, much more than Numi. JDV is like the Chuck Norris of finance. He eats credit crunch for breakfast.
i feel that trends represent momentum and EMH represents those of us who still have jobs.
"That paper mentioned above is just the “markets were wrong therefore they weren’t efficient” nonsense. Stuff like “The academic world provides a basic theorectical understanding of the markets, but in practice it is much more dynamic” is nonsense. " Interesting comment. I’m well aware of the publish or perish life that academics are involved in, and that as a researcher they are mostly out to get cited by the peers. However I’m not aware of how that has anything to do with this thread. The paper referenced in your post states: “Undeniably, the highest expected return stocks are, collectively, highly attractive; the lowest expected return stocks are very scary – results fatal to the efficient market hypothesis. While this evidence is consistent with risk loving in the cross-section, we also present strong evidence consistent with risk aversion in the market aggregate’s longitudinal behavior. These behaviors cannot simultaneously exist in an efficient market.” and provides evidence to back up this claim. And yet you offer none. I’m pretty sure that’s called a straw man logical fallacy.