Random Walk and Outperforming Fund Managers

I find it interesting that so many of these people are Trend followers ….. and don’t follow any fundamentals whatsoever ……. what does that tell us about fundamental analysis and the many CFA’s walking around

I think it’s better to allow them to continue being deluded, Barry. The fewer people who use technical analysis, the better. Let ’em continue to think it’s worthless. We’re better off.

Correction, I meant “perceived efficiency” instead of “perceived inefficiency”

I trade the German Bund futures for a living. I am based in Asia and I know “nothing” about the German and Euro economic conditions. I don’t even follow ECB monetary policies.

However, what I do know is the price action of the bund – all the intra day high and lows (I know them by heart to read the tape) and I know the daily chart price action.

That’s all I need to know to trade – no fundamentals whatsoever.

It’s the simplest and most effective way to trade.

Now that the efficient market hypothesis (and capitalization-weighted index funds based on it) is discredited as bogus, the latest and greatest fashion in the indexing world is “fundamental indexing” as described here:

http://www.2000wave.com/article.asp?id=mwo061606

And as sold to the public here:

http://wisdomtree.com

Barry has discussed this in the past, but not today. How does fundamental indexing alter the anaysis, if at all?

(Also, I am still waiting to hear what Barry wanted to say on Kudlow about housing last night.)

“If technicals are voodoo, and quantitative data irrelevant…”

Straw man alert.

I’m with Joseph Ellis (author of Ahead of the Curve) — to have a good predictive technique, you need a plausible hypothesis supported by data. A purely quantitative approach, e.g. predicting the stock market on the basis of who wins the Super Bowl, isn’t going to work in the long run…

I’m sure Warren Buffett , Marty Whitman , Tweedy Browne , Chris Davis , Roy Neuberger , Leucadia use technical analysis !!!!!!!!!!!

Perhaps Larry Summers said it best (after the 87 crash): “The efficient market hypothesis is the most remarkable error in the history of economic theory.”

Technical and fundamental approaches can both work wonders in the right hands. And of course they can work spectacularly well in concert. Surprisingly, many virtuosos of one school don’t recognize the value of the other. (Soros is on record as saying technicals don’t work. We know how the trend followers feel about fundamentals.) When it comes to investing approaches, there is an odd provincialism that borders on the religious. Maybe it’s the tribal tendencies in our genes.

In my opinion the failings of random walk theory are symptomatic of neoclassical economics’ biggest problem in general: a tendency towards gross oversimplification. Economics started out as conjecture and soft theory–guys arguing with each other subjectively and off the cuff, the way poets and philosophers do. The neoclassical approach rescued economics from this fuzzy-headed swamp by applying rigor and logic, but they got so excited about the trappings of hard science that they threw the baby out with the bathwater. “Man, humans are messy. Wouldn’t it be cool if we could invent a robot called ‘rational economic man’ and study him instead? Wouldn’t the formulas be so much more logical and rigorous that way?” And so they took that ball and ran waaaaay too far, creating an artificial world that resembled reality in some ways but grossly distorted it in others. As if in their zeal for exactitude, the neoclassical economists decided that Pi could just be rounded off to 3, and all irrational numbers could be ignored. There are plenty of other areas besides random walk where gross oversimplification dramatically underserves–trade deficit calculations for example–but random walk is the most famous example because it got turned into a rallying cry for mediocrity (hooray, we’re all average!) and a ubiquitous index-fund marketing tool.

The other reason random walk still survives (in my opinion) is because economics as a field is so polluted by corporate and political influences, and the random walk idea generally serves those influences. The reason that the vast majority of managers can’t beat the market is because real investing / trading skill is rare and thus carries a high premium, but that truth is horribly non-egalitarian and doesn’t play well in Peoria. Better to say no one is skilled than admit that rewards can and will accrue to a small minority on a consistent basis.

Not to mention that the vast majority of economic forecasters and prognosticators have personal biases akin to cancer researchers getting funded by big tobacco. On top of that, wishful thinking runs rampant in economics and politics–and the myth of a level playing field appeals to politicians, idealists and Wall Street marketers alike.

Given a wide number of market participants, outperforming the market over the long term is proof of nothing and is easily captured in a random scenario.
Consider the exercise I remember from finance class: Every student (50 or so) in the room is given a coin at “year one” we flip. Heads, you beat the market, tails, you don’t and sit down (normally half the room). Next round, “year two” same thing, another half the room sits down. Fast forward to “year 10” and a few “market geniuses” have beat the market for 10 straight years…oh my god, how did they do it…

You can debate until the cows come home whether efficient market theory holds or not. One thing you cannot debate is the performance of index funds. They routinely outperform actively managed funds. Before all you stock pickers fly off the handle, fund managers can revel in the fact that BEFORE fees (mgt, transaction costs, taxes) the average fund manager does outperform their relative benchmark. So, what is the lesson learned? Fund managers prove the market efficiency theory wrong every day (can you say Seth Klarman?). However, they are not good enough to consistently outperform their benchmarks and the fact that efficient market theory is about 80% correct.

For you non believers check out Dimensional Fund Advisors (DFA). They do not pick stocks. How do they consistently outperform their benchmarks? They believe in efficient market theory and they operate as a block trading partner for counterparties that want to dump “bad” stocks. Because of this clearing function that they provide and the credibility that they have earned, DFA almost always acquires stocks a 50-150 basis points below the current market value. They don’t spend a dime on research and they trade so efficiently that they basically eliminate transaction costs. They run a cheap operation and bet on high book/market ratios and the efficient market theory.

The take away is that you can find plenty of “smart” managers out there. I say “smart” because equating skill to outperformance is difficult. More often than not the outperformance is luck. Nevertheless, most of these “smart” managers cannot beat the flawed efficient market theory when they gouge you on fees.

BR: Problem is, index funds have underperformed the broader markets for the past 6 years . . .

X’s argument above is made repeatedly in the book Fooled by Randomness. You set enough monkeys typing and someone produces the encyclopedia britannica sooner or later. There are an awful lot of monkeys issuing buy and sell orders to the financial markets.

That isn’t to say that the efficient markets hypothesis is true. It just means the existence of winners with well-defined trading strategies isn’t a counter example. From X’s example above, the student who won probably had a very good song and dance about why he predicted the coin flips correctly.

I’m interested in understanding this better. I believe the random walk proponents believe that even if you beat the markets for many years, you cannot do so consistently. The DFA guys even have a list showing market timers underperforming the S&P500:
http://www.ifa.com/Book/Book_pdf/04_Time_Pickers.pdf

Fund managers prove the market efficiency theory wrong every day (can you say Seth Klarman?). However, they are not good enough to consistently outperform their benchmarks and the fact that efficient market theory is about 80% correct.

Here’s what I don’t get. Who needs a theory that is 80% correct? Why not just throw the $@#$# thing out and admit it was close but not cigar?

Imagine if engineers were taught structural load theories that were 80% correct. How well would that work?

There is nothing wrong with “I don’t know,” or “We don’t have a full answer yet.” Plenty of serious scientists make these statements all the time in other demanding and rigorous fields. Yet here we’re supposed to say “Screw it, EMH is clearly jacked up in significant ways but it gets kinda close so we’ll go with it???” That is an intellectual travesty.

p.s. I don’t see the controversy in pointing out that index funds routinely outperform actively managed funds on balance. This is not an argument in favor of efficient market theory at all! It is an argument in favor of different assertions: that less activity is better than more when it comes to large sums, perhaps, that decisions of general allocation and exposure are more important than decisions regarding the manager selected, and, most importantly, that alpha is in limited supply and will only be generated by a minority of exceptional performers.

But once again, the random walkers work backwards here. They say “the Seth Klarman instances are few, so we can ignore them or pretend they are just anomalies.” A truly scientific approach, that seeks to explain the phenomenon in question instead of seeking to conform evidence to a preconceived notion, would NOT ignore the Klarman instances, however few, that blow the theory all to hell. Instead, honest economists would willingly go back to the drawing board, and engage in more research and reduction, instead of clinging to a too-big, too-general, too-goofy, “80% correct” theory. But this isn’t happening because inertia, politics, incentives, and ego all play an entrenching role.

Trader75,

Compare the fees you can get for an index fund and that which you get for a mutual fund, and I think it’s pretty obvious why Wall Street isn’t too enthused about index funds. ETFs are becoming a different story for brokerage houses, but only because people will trade them like stocks.

“how crazy is the idea that you can just throw your money into some passively managed vehicle . . . and X years or decades later it will all have worked out swell for you.”

Not crazy at all, if you understand the concepts involved, the importance of allocation, and the aspects of behaviorial investing Barry frequently references such as overconfidence and a misapprehension of risks by investors, and look at performance histories. VFINX’s record over the past 30 years is 12% annualized. The percentage of investors, pro and amateur, who’ve had a run that exceeds that is pretty small.

Wow. You sure are fired up about EMT. I take it from your well-stylized prose and heady vocabulary that you fancy yourself a pretty sharp guy.

Actually no ML, not all that fired up. Just enjoy a good debate. It’s a fun and interesting topic, although not all that important in the big scheme of things, as others have pointed out.

I fully agree with you that the majority of active managers are, if not necessarily crooks, at least not worth their fees. I used ‘alpha’ in the offhand sense to refer to outperformance vis a vis a broader group of managers; of course you are right that, at the end of the day, absolute performance is what matters. (Which, by the way, doesn’t strike me as an EMT-friendly belief. Hmm.)

Not sure what to do with the sharp guy comment. I write how I write. Don’t have a dictionary by my side; am I to be punished for turning a phrase? (Insert shrug and puzzled look.)

Royce, fair point on VFINX and the like. In no way did I intend to dispute the truth of that observation, that some passive vehicles have had excellent past performance. I just happen to agree with Larry Summers, that EMT is a remarkable error with a lot of interesting causes.

As to the outperformance of certain vehicles and measurements in US markets over the past hundred years, I would argue that is primarily tied to the shining performance of the 20th century US economy, which in turn is better explained by big picture historic / economic / cultural / geopolitical factors than an overgeneralized economic theory such as EMT. And while we can look back and see 30 years periods of “Woohoo!,” let us not forget there are also multi-decade periods of “oh crap!” in which an active strategy (or even a sideline strategy) would have beat the pants off a passive one.

There are certain things about EMT that make sense and certain conclusions born of EMT that are wholly logical. But on balance, I think the whole exercise is so flawed that we would be better off salvaging what’s good from EMT, scrapping the name, and calling it something else entirely. Or better yet, dispensing with the need for a grand theory of markets in the first place. Physics hasn’t come up with a grand unified theory yet, and particles are nowhere near as messy as human interaction, so why should markets have one? I’m mostly against oversimplification I suppose.

Cheers

Gents/Ladies–wish i could join this lively debate, but after a 30 minute thoughtful [or so I thought] piece i put together–I was “spammed” blocked by the sites [?new] software.

Wondering if others are having this problem? It’s cool to call Kudlow a “drug addict” and other in sundry ad hominem posts I’ve see lately, I guess, but thoughtful essays are blocked…

I sent my piece directly to BR—so “what say ye” sir?

Hope it’s a software glitch, otherwise, the value of this site/discourse, for me, decreases by an order of magnitude.

RB, to understand it better you really need to read someone like Malkiel rather than focus on how random walk theory is characterized here. The basic argument is statistical: it does not posit that the market is actually random but that stock prices can be successfully modeled as if it were. Nor does it posit that statistical outliers (outperformers) can not exist, only that there is no way to guarantee such performance can continue.

In its weak form it posits that the history of stock price movements contains no useful information that, on average, will enable an investor to consistently outperform a buy-and-hold strategy in managing a portfolio; e.g., a lot of folks thought Granville was the ace, then when he fell it was Precter, than Garzarelli, etc. so how, on average, would anyone know they found real talent in an advisor or had the talent themselves? It’s a matter of statistical uncertainty — on average they can’t and/or won’t — but it’s not a matter of impossibility even if some academicians appear to say so.

The strong form of the theory is quite beyond the pale of sensibility and even Malkiel mocks it a bit, offering the example of an economics professor who sees a ten dollar bill on the sidewalk and passes it by, saying to his students who are attentively trotting behind, “ignore that bill on the sidewalk because if it was really there someone would have already picked it up.”

Our host has posted a number of comments from successful traders regarding what makes for success and many posters here are also successful and make knowledgeable comments in turn (sometimes casting aspersions on the comments previously posted – lol). Clearly long-term success is possible so what’s really at issue?

Some years ago I did a study of the Pacific Stock exchange incorporating both quantitative and qualitative (ethnographic) techniques. Made some good acquaintances at the trading desks in the process and gained some insight into trading success that, boiled down to it’s essentials (and with operational skill as an already given), could be described as (1) superb situational attention and (2) excellent money/risk management. Too much detail for a post but the bund trader who posted above gives some insight into the former quality.

The problem being that, according to the weak form of random walk, no investor on average can rely on any of that over the long term. I think Harry Browne says it better though (I paraphrase): Really successful investors are like exceptional athletes or exceptional people in any field, they have a talent, and exceptional talent by definition is not common or we wouldn’t recognize it as such. Unless you (the individual investor) have clear evidence you possess comparable talent you had better learn how to diversify, manage costs, and trim the odds against you.

well put RW–

‘selection bias’ is the operative statistical term and needs to be clearly understood by all who make money manager selection decisions [or choose not to…]. That, and ‘regression to the mean’, are arguably the most important ‘concepts’ for someone trying to manage their own money. And, in my opinion, need to be deeply understood.

This has been a valued site in many ways, but the number of folks with what I call “guru-itis” is astounding. There are now over 8,000 hedge funds, not to mention thousands of other actively managed non-hedge vehicles, but 20 names are proof of the ability to confound the market? I have learned a lot from this site–specifically the piece on markets following p/e expansion and contraction not earnings as hugely enlightening. Thanks to BR. But I’m beginning to wonder how one can justify on the one hand telling everyone that forecasting is foolish, yet later tout the genius of those [few] who made millions doing it. After all, there’s a one in a thousand chance of getting 10 heads or 10 tails in a row–if 10,000 people play the game, ya got your 10 guru’s.[retrospectively of course–which is the point]

No, I’m not Taleb with a pseudo-pen name, but consider reading ‘Fooled by Randomness” folks. Consider understanding in depth the two most important statistical factors at work in trading–selection bias and regression to the mean.

Yes, I called the real-estate bubble, emerging market bubble, and small cap bubble–but, alas also got out earlier than I should have to have created maximal profitability. Recognizing market imbalances actually isn’t too difficult if you can control primal ‘greed and fear’ instincts, but that doesn’t mean you can nail the top and bottom with any certainty. And when you add in fees, and most importantly taxes—a topic I have yet to see discussed here [namely 15% long term cap gain hit v. what I suspect is 40-50% marginal rate for many high income, high tax state folks here for short trades– is huge], the value add of being active becomes harder. [abolish all profits on capital gains, and we might have a different game…]. Yet we aare here trying.

Lastly, in case there is a young PhD in economics candidate out there, what about developing and proving a ‘Universal Field Theory of Efficient Markets”? Namely, perhaps that markets ARE truly efficient, but to BR and the other critics–perhaps what you are objecting to is that those theorists are not in fact measuring ALL the component variables. So, if one could accurately measure market fundamentals, technicals, sentiment, neuroeconomic decision making, macroeconomic/social- political volatility, one may actually be able to say the market is, in a sense efficient in that all data is being measured and reflected in a price instantaneously; –very different than saying that an individual may agree/disagree with that calculation. And, as far as I know, this is not capable of being done anyway—but may make a nice doctoral thesis though…

One last rant, as a [successful] dot.com entrepreneur that sold his company in 2001, one comment that people forget in the current rise to glory of Google et al. Guess what, the ‘eye balls’ metric that the VC’s used to justify valuations in the 90’s—which was first embraced [overly so with typical human greed], and then sneered at after the crash [also typically, due to human fear] was actually [news flash–heard it here first] right! It was, in fact, the correct metric to use—it took time and Google, [e.g. Google’s adSense] among others to figure out a way to monetize the phenomena, and regression to the mean to bring growth prospects of Internet commerce in line with reality. MySpace just sold for $500M because it was able to draw 50M unique visitor in under a year, advertisers have an easy way to measure ROI for their investment, and online marketing works.

The facts are, the greatest way to amass great wealth quickly is to 1. inherit it [sadly, not my case–but probably screws a fellow up anyway…], or create a great company. Investing is key, but unless one is actually gambling, is a longer, slower process.

Thoughtful comments/responses encouraged–but please no flames. I tried [and it was my earnest goal] to be respectful to all. Just needed to respond to this trail.

Cheers

Index funds have only underperformed if you’ve been in the wrong index funds. Emerging markets, oil, broker dealer, transports, small caps, energy, metals, etc, have all outperformed as indices.

Trader75, how many iced venti mochas have you had today? You went on a rant proportional to mine. lol.

Btw, “Imagine if engineers were taught structural load theories that were 80% correct. How well would that work?”

They were. It’s called the Big Dig in Boston. lol.

The way I see it, it’s not so much an excess of outperformers as an excess of consistent outperformers. Dart-throwing monkeys will fall off a cliff at a certain rate; if you run some plausible numbers, you will see that the stars do fall off the cliff, but not at a high enough rate to make any sense at all.

In re: DFA, if they beat a large-cap index by 150bps without taking on any relative risk, I’d like to see that documented. In small-caps, I absolutely believe their methodology works.

My advice for generating alpha to those who don’t get paid to do it directly is to index substantially all their money and to get a raise. When you’re 65 this is bad advice, I grant, but I chat with my peer group, and at 30 or even 40 a 10% raise usually beats 100 bps of alpha, and that is indeed without taking any relative risk.

Barry,

You of all people should know that data sets always contain anomalies. Out of literally millions upon millions of data points you point to a handful that are near the higher end of the range. In stats, these points prove nothing. You must look at the entire set to prove your point. If you throw 1,000 darts and hit bulls eye 50 times it is not fair to say that you hit bulls eye 50 out of 50 times. These throws prove very little when looked at in the entire data set.

Money managers are the same. A small handful actually outperform a standard index fund over the very long term. This is even more true for hedge funds. I would die to see how some of those funds do when you back out taxes and fees. 25% for fees and 35% for taxes is why most hedge funds are a horrible investment.

This is not to say that outperforming the market is impossible. It just goes to show how extraordinarily difficult it is. Most people are better off saving themself the trouble.

It’s funny to hear so many traders trash EMT on hand and arbitrage on the other. Arbitrage only holds under EMT.

It’s easy to run out the tired old LTCM example, but in doing so, one ignores the billions of dollars those guys made for 4 years. Their failure was a function of not understanding the principles of correlation under market panic. Russia defaulting on bonds did not equal a correlation of 1 in LTCMs models. That mistake sunk them. All they money they made previously was pure EMT at work.

Two points.

First, on Vanguard, specifically:

Vanguard’s small cap fund, for example, has done better than 90% of managed funds in its category over the past 5 years, even though its manager does not use technical analysis and all the various tools you’ve mentioned

Yes they do. Vanguard, Barclays Global Investors and State Street are probably the biggest users of quantitative analysis (and certainly the biggest employers of quants for investment management) in the market.

The reason for this is that Vanguard et al do not, in fact, follow the strategy of buying and holding the entire index. They don’t because this is a monstrously inefficient trading strategy for a firm that has to provide liquidity for inflows and outflows – the average position size is well below the optimum.

What index trackers do is to construct a portfolio which minimises tracking error. There are many such potential portfolios, and a considerable degree of skill goes into constructing the best one. Also, when constructing your tracking portfolio, you are always looking to give it an “alpha tilt” – a little bit of outperformance to help pay your management fee. Vanguard in particular are gods at achieving this.

What people need to realise is that there are good and bad trackers (some of the worst-managed tracker funds have performance that is nothing short of embarrassing) and if you are comparing active managers to Vanguard, you are comparing them to the Warren Buffet of trackers.

Second, on the coin-tossing issue:

This isn’t a valid statistical argument. It is true that if you hold a coin-tossing tournament, somebody will win. But if you then hold a second coin-tossing tournament and the same guy wins, then have a look at his coin because this would be an extremely unlikely event under the assumption of a fair coin.

Buffet, Soros et al were identified as great investors at least twenty years ago. At that time, it might have made sense to call them “Lucky tossers”. However, they have continued to maintain their performance records ever since then. The hypothesis of an unbiased coin has hardly any support at all.

I dont’ know if this thread is done, it’s a pretty good one and a frequent topic of conversation. If there was only one person who consistently outperformed the market, then most people would agree it could be luck. If there were thousands most people would agree it would be skill. I’ve heard people use the term semi-strong EMT, which may be better. Trying to strickly apply EMT would be tough.

But, it also begs a lot of philosphical questions in defining alpha itself. If you say you outperformed the market, then what market are you talking about? If you were in US stocks while the market in Finland went up 40% or out of NASDAQ when it went up 30% (or whatever high %’s it was hitting), did you underperform? Time horizon and asset base is very tricky and the framing is difficult.

I think the economist did a little piece once showing how much money you would have if you just put $1 into the best market each year world wide over the last 100 years. You would buy Jan 1 and sell Dec 31. The result was something huge, I think they said more than all total world assets.

“Because you said that Vanguard doesn’t pick stocks or use technical analysis and they do.”

Their goal is to model the index involved, not identify which stocks will outperform other ones or predict the movements of markets using technical analysis. So they use TA in their model. So what? You’re bringing up an irrelevant detail on an issue nobody is discussing.

Speaking of intellectual error — “These outperformers include many technical and quantitative driven strategies.” The “outperformers” you listed were people, not strategies. It does not necessarily follow that the strategy used by the investor is the ultimate source of the investors success. Trade75 argues that real investing skill is rare (backing the claim with pure twaddle – “it’s because the public won’t tolerate excellence like mine!” – please) is consistent with trading success not being well understood. It is no accident that techies mock fundies and quants mock both. Each group, wrapped in the normal human urges for affiliation and control, claims that it is the trading theory practiced by members of the group that account for their excellence. Iconoclasts (whose affiliation is with other iconoclasts) claim that they use a bit of this and a bit of that, but still claim that they have “knowledge” which they may not, in fact, have. If all this stuff works, some of the time for some people but not all of the time for anybody, then still we don’t know what works.

Mighty cheeky to crow over Malkiel and Bainbridge committing supposed intellectual lapses and waving around behavioral economics as evidence, in the midst of one’s own intellectual lapse, based on some common forms of irrationality.

Outperforming fund managers

So they use TA in their model. So what? You’re bringing up an irrelevant detail on an issue nobody is discussing

If the fact that they do is “an irrelevant detail”, then why did you, in your original comment, think it was so important to say that they didn’t?

The quantitative strategies used at Vanguard and other trackers have two objectives; to minimise tracking error, and to generate alpha. “Generating alpha” is done by picking those stocks which will outperform the others.