Investing Myths: Alpha and Beta

[NajdorfDefense]

This will be a complex, multi-part tl;dr, so be warned.
Short form: Buy beta cheap, but search for Alpha.


Investing Myths: Alpha and Beta


There is a lot of confusion and misinformation on these boards about investing and the markets in general, so I thought I’d try to clear a few things up, using a somewhat advanced discussion in places and the basics in others.

Myth #1 – No one can beat the market, so you should put all your money in a Vanguard equity index fund. Strong-form market efficiency theorists say that NO ONE can ever beat the market over time, or show skill superior to a dart-throwing monkey.

First off, putting all of your money in equities is a huge mistake for 90% of the investing population. Only the very young should even consider doing so, and not use any cash that could be needed in the near future.

This myth is biggest, most standard myth around which most other investing myths are built. It is a complete lie, of course. Think of investing as very akin to a poker game, you are playing against the other players, and there is a cost, or rake, to being in the game. Can you beat your local NL home game? Can Doyle or Ivey? Of course. Can everyone be above-average? No.
Superior performance is what we call ‘Alpha.’

Many investors have demonstrated the ability to beat the market over statistically significant periods of time with superior performance – Warren Buffett would be the most famous example, John Neff’s Windsor mutual fund crushed the market over 32+ years, beating the SP500/SPX by 3.1% annually, after fees. $100k turned into $2.5mm in ‘the market’ but $6.1mm in Windsor.
James Simons has grossed 50% annually for the last 20 years [longer, if you use his personal results prior to managing outside money.] Steve Cohen is coming up on 15 years of 50% returns.

While there are a variety of ways to beat the market, from Michael Marcus’ technical trading, to Simons’ statistical arbitrage, to Sussman’s multi-strategy, to Thorp’s bond arbitrage [who wrote Beat the Dealer] or Sandell's, the majority of professional investors who beat the market are value-disciples from the school of Graham and Dodd, again with Buffett being the most famous student of theirs, having done it for 4 decades.

They seek to buy undervalued, out-of-favor stocks with high cash flows, low book values, often low debt, that generally pay some sort of dividend. A simple example would be Neff’s purchase of Ford in 1984 when it carried a PE of 2.5x. In less than 3 years it climbed 350%. Strong-form says the market is 100% rational, 100% of the time when evaluating every single security in the marketplace -- this is clearly false.

Myth #2 – You can’t predict which managers will outperform.

They insist you buy your equity exposure via a relatively passive index, pay Vanguard’s moderately expensive indexing fees [when compared to BGI or SSgA] to purchase a market-weighted index, and guarantee that you underperform the SPX or Wilshire 5000. You guarantee yourself sub-market returns in perpetuity, but at least you’ll get a relative return that is close to ‘the market.’ Even if that return is -10% a year for the next decade. This is what we call ‘Beta’ exposure.

However, if you invest Other People’s Money for a living, will they be happy at the end of they year when you tell them, ‘Well, the market fell 20%, most funds were down 22%, but I was only down 20.2%! [plus my fee]’?

They’ll be unhappy and you’ll be fired.
Individuals care about absolute returns, not relative returns. Only institutional investors who invest in every sector of the stock market [should] care about relative returns.

Indexing doesn’t guarantee you a positive return, much less an inflation-beating return. Remember, if a 20-yr track record of beating the market doesn’t guarantee superior returns, neither does the history of index investing. The only guaranteed returns come on US Treasury bonds, GNMAs, and bank deposits up to $100k, that’s it. As of end of March 2007, the SP500 returned about 29% gross, before fees, in 8 years -- do you consider this good performance? It trailed any reasonable asset allocation model as well as bond portfolios.

Once you demonstrate the absurdity of Myth #1, the indexing die-hards will retort that just because Buffett has beaten the market for 40 years, or Richie Freeman’s mutual fund for 24 years, doesn’t guarantee Buffett will do it for the next 40 years. And that it is impossible to predict who will.

Well, Warren won’t live to 115, and while it is obvious that no one can guarantee Alpha on an annual basis – value investors will freely admit there are times when value suffers relative to growth, which allows them to buy more of their favorite stocks cheaply – you certainly can predict managers that have a consistency of outperformance will continue to beat the market.

Academics have proven this time and time again:
Ibbotson and Chen's academic paper on the topic of Hedge Funds--http://papers.ssrn.com/sol3/papers.cfm?abstract_id=733264
using Ibbotson's data for equity L/S funds only [Jan 1995- April 2006:
'L/S Hedge Funds create a net alpha of 5.41 per year - that is net, and overall return of 13.10%.

All results are for dead + live funds, and do NOT include backfill data/bias. Alpha per HF strategy:
Equity Neutral +1.94%
Fixed Inc Arb +3.91%
L/S Equity +5.41%.'

add'tl- Goetzmann and Ibbotson's 1994 study, 'Do Winners Repeat?' Journal of P.M. and the updated ‘Do Winners Repeat with Style?’
ROGER G. IBBOTSON , AMITA K. PATEL February 2002
http://papers.ssrn.com/sol3/papers.c...ract_id=292866
'Several studies have found that considerable persistence exists in mutual fund performance. We study this phenomenon in fund managers who achieve superior performance, after adjusting for the investment style of the fund. Our data of domestic equity mutual funds indicates that winning funds do repeat good performance. Style-adjusted alphas are evaluated on both an absolute and relative basis.'


Alphas, Betas, and Hedge Funds

Now, those pounding the table in favor of indexing via BGI/SSgA/VGuard have a point: It is a very cheap way to access Beta exposure – which is simply exposure to the index you are trying to replicate. This should be very low cost – you may have heard of some funds being called ‘index funds in disguise,’ that is they charge you 1.25% and own 90% of the stocks in the SP500 – that is truly a waste of your money.

By definition, Alpha – or market outperformance – has to be a net Zero for all investors, actually slightly lower including fees/trading costs – just like a poker game.

Indexers will tell you that since a majority of mutual funds underperform ‘the market’ – not 90% as they claim, but many – that this proves you should just index. Burton Malkiel says that mutual funds regularly underperform by 2% on an annual basis. Since the average mutual fund costs about 1.3%, that means that most funds are losing 70bps per year in alpha. Index funds are losing 15-20bps per year in alpha.
However, this also means that other players in the market, by definition, are generating significantly positive alpha. After all, if 70% of funds have negative alpha the other 30% are gaining all of that alpha, plus the alpha opportunities ignored by the enormous cap-weighted index funds like the SP500 or Total Market Indexes. [Siegel and others have invented value-weighted indexes to take alpha from cap-weighted ones.]

In less than 3 years from 2000-2002, the SPX lost 45%. Over the entire period, it lost 40%. Beta/Index investors got the relative return they were looking for!
I’m sure many of you were investing back then – how did your index funds do?

If you can identify superior managers, something that may not be easy or trivial to do [but many sites like MStar, Lipper, etc can give you the foundation] you can certainly do better than the market AND with less volatility, given time and a modicum of work and intelligence.

Myth #3 – Risk and Return are directly related.

I can see your eyes rolling now:
“Of course they are, they have to be, everyone knows risky assets must generate a higher return and vice versa!’

This is standard CAPM theory, however, Fama and French last year said CAPM had virtually no empirical support, updating a study from 15 years ago. Most economic models assume a positive-variance return related directly to a risk-aversion coefficient. Diversification is a free lunch, and too little exposure is as bad as too much exposure to an asset.

For decades, researches have looked for evidence in the equity markets of a linear relationship of return and risk using market indices and realized returns. Many now accept there is simply no correlation as a practical matter. Whether the academics modeled risk as varying with volatility, cash flows and discount rates, consumption risk, risk and hedge components, CAPM is a failure.

Both Lo and Black were skeptical of the ‘size factor’ as an effect on risk, due to the lack of any theoretical reason for riskiness. Fama [1993] suggested both size and book/market represent distress risk, which is harder to measure as it has a power-law tendency to jump to zero, and may be linked to difficult-to-measure brainpower at a firm [as firm declines, good employees leave, vicious circle].

Several firms have used models and bond agency ratings to predict defaults to evaluate the equity performance of various ‘risk’ classes. Generally speaking, they find distressed stocks have abnormally low returns, inconsistent with return/risk assumptions. These firms have higher volatility, betas, and market cap-factors than stocks with a low risk of failure.
Studies have shown that stocks with low risk factors such as lower beta, leverage, higher profits and dividends outperform the market, consistent with the research on distressed firms. O'Shaughnessy shows similar results.

Others have tried to measure ‘risk’ as uncertainty, and failed. Researches have shown it is not positively related to equity returns, and may even be negatively correlated.

Corporate bonds are no better – recent yields on the High-Yield index were 270bps better than BBB-rated corps, yet the default rate was 3.5% worse according to Moody’s, suggesting a negative reward for investing in risky debt. [Falkenstein discusses all the above in more detail]

This is why breathless news articles talking about ‘risky’ Hedge Funds always make me laugh – the authors generally have no clue what they are talking about. Sure, Amaranth went out of business, but how many US firms also did last year? Worldcom, Enron, Parmalat, et al, all were 10-20x larger than Amaranth, and Amaranth wasn’t the sole retirement/401k vehicle for any of its investors. Fannie Mae fraudulently created, and then vanished $8bn in earnings, but CEO Franklin Raines got $tens of millions in bonuses and is a huge fundraiser on Capitol Hill.

Superior Alpha thru Hedge Fund Investing:
I decided to go back and look at the portfolio of hedge funds my clients are invested in, and see how they did during the downturn of 2000-2002. This is ex ante, not ex-post data, these are actual investments made before/during those years, not cherry-picked, all returns after fees, of course.

The SPX dropped 40% over that time frame, with a downswing of 45%.

The basket of 9-11 hedge funds rose 51% over that time frame, with a downswing of only 5%. It outperformed 26 of 36 months. It had only 5 down months,of minus 4.49% total, versus an SPX drop of 13.9%.

Monthly SPX Basket SPX>? Month SPX Return Basket Return
Dec 02 (6.03) 1.83 0.00 12/01/02 0.60 1.51
5.71 0.96 1.00 11/01/02 0.64 1.48
8.64 2.32 1.00 10/01/02 0.60 1.47
(11.00) 1.71 0.00 09/01/02 0.55 1.44
0.49 0.50 0.00 08/01/02 0.62 1.41
(7.90) (1.36) 0.00 07/01/02 0.62 1.41
(7.25) (2.13) 0.00 06/01/02 0.67 1.42
(0.91) 1.07 0.00 05/01/02 0.73 1.46
(6.14) 0.36 0.00 04/01/02 0.73 1.44
3.67 0.27 1.00 03/01/02 0.78 1.44
(2.08) 1.08 0.00 02/01/02 0.75 1.43
(1.56) 2.13 0.00 01/01/02 0.77 1.42
Dec 01 0.76 2.33 0.00 12/01/01 0.78 1.39
7.52 2.32 1.00 11/01/01 0.78 1.36
1.81 1.40 1.00 10/01/01 0.72 1.32
(8.17) 0.69 0.00 09/01/01 0.71 1.31
(6.41) 1.70 0.00 08/01/01 0.77 1.30
(1.08) (0.45) 0.00 07/01/01 0.82 1.28
(2.50) 1.00 0.00 06/01/01 0.83 1.28
0.51 2.42 0.00 05/01/01 0.85 1.27
7.68 (0.15) 1.00 04/01/01 0.85 1.24
(6.42) 0.80 0.00 03/01/01 0.79 1.24
(9.23) 0.74 0.00 02/01/01 0.84 1.23
3.46 2.12 1.00 01/01/01 0.93 1.22
Dec 2000 0.41 1.47 0.00 12/01/00 0.90 1.20
(8.01) 3.78 0.00 11/01/00 0.89 1.18
(0.49) 1.73 0.00 10/01/00 0.97 1.14
(5.35) (0.39) 0.00 09/01/00 0.98 1.12
6.07 1.03 1.00 08/01/00 1.03 1.12
(1.63) 1.73 0.00 07/01/00 0.97 1.11
2.39 1.34 1.00 06/01/00 0.99 1.09
(2.19) 2.93 0.00 05/01/00 0.97 1.08
(3.08) 1.37 0.00 04/01/00 0.99 1.05
9.67 0.83 1.00 03/01/00 1.02 1.03
(2.01) 0.55 0.00 02/01/00 0.93 1.02
Jan 2000 (5.09) 1.74 0.00 01/01/00 0.95 1.02
Average month (1.27) 1.16
Net (40.10) 51.07

As you can see, the basket returned 2.52x the SPX in just 3 years. Amazing. This is what’s known as having a lot of alpha in your portfolio, and with much lower volatility as well. Every fund in the basket was up during that period, not one had a negative return, and correlation was less than 0.2.

Now, I’m not saying Hedge Funds will always outperform, in strong up markets, they will almost always lag due to being less than 100% long, they are hedging, after all. Looking at last year, the basket returned almost 14%, so about 1% behind the SPX. Sadly, due to our ‘geniuses’ in Congress, most people will never be able to access this investing talent, unless you are lucky enough to work at a bank/state that somehow has finagled part of their retirement plan into one of these funds.

Myth 4: You can’t get Alpha from Beta.

This seems obvious and deeply true on its face. You can buy beta, or sell beta, but that doesn’t create alpha, or much of chance to get it assuming that any temporary dislocations created from you buying more of the index are either absurdly small and/or temporary and immediately taken by the trading desk doing your executions.

However, there are market-neutral strategies that use Beta Arbitrage to exploit pricing errors in the market, as discussed by Black, and Vuolteenaho, among others.
To get pure Beta exposure, you would buy the SPX and sell bonds. Another way is to buy high beta stocks and short low-beta stocks. If these two ways to obtain the same risk earn different returns, there exists an arbitrage.

Fischer Black suggested a tilt towards low-beta stocks and away from high-beta. This could be a constant allocation regardless of environment. Buy $1 worth of low-beta [0.5] stocks and sell $0.33 of high-beta [1.5] stocks, resulting in a beta-neutral position.
If the return over risk-free rate of high-beta is less than three times the return on low-beta stocks, you made an arbitrage profit, and vice-versa.

What creates these low-beta arbitrage opportunities? Black, Jensen and Scholes suggested investors want a hi-risk, hi-return portfolio but are shy of leverage. Karceski suggests long-only managers prefer hi-beta stocks because investors chase performance. Money flows are asymmetrical, in other words, and managers must do well in hot markets or see outflows, whereas in bad markets people do not blame their manager. Finally, something we can all agree on, irrational traders, day and otherwise, causes overvaluation and high betas, many of whom won’t leverage up either.

But is this really a value strategy in disguise? Daniel and Titman [97] and Davis, Fama, French [2000] show otherwise, by using stocks with the same P/B ratios divided into quintiles, and a large historical premium for Beta Arbitrage appears.

Tactical Beta Arbitrage can be employed instead, when equity premiums are high and betas are low, as well as the reverse, which implies trading opportunities. Naturally, you’d need a good equity premium model to do this. During periods of strong momentum for high-beta stocks, low-beta stocks are less attractive when the market also forecasts a low equity premium for stocks. Again, this would be a tactical bet, not a permanent one. [cf. Vuolteenaho]

Finally, another way to do the above if you don’t believe any manager can systemically create alpha, is to buy Beta thru BGI/SSgA/anywhere, and short Hedge Funds who have much higher fees and expenses. Buying the Beta is cheap and easy, and in London, Luxembourg, and many other markets you can short Hedge Funds from Man, Soros, and dozens of others. If, over time, they underperform by 2%+ as Malkiel suggests, you will make 1.8-1.9% alpha/ profits annually on a market-neutral position, [which you could certainly leverage], using a basket of funds to bet against to minimize your volatility of returns.

[ConstantineX]

This is really excellent. I read similar ideas on Brad de Long's blog. However, he suggest that middling hedge fund managers are overcompensated. Read: "Not at the top, further down the food chain, however... it is a mystery how the hedge funds staffed by very smart and hard-working people who nevertheless do not seem to have much of a risk-adjusted edge over the market indices nevertheless collect fees of 2% of assets and 20% of returns each year..." and "For a generation, investors in Fidelity funds received net annual returns of S&P - 2.5% + noise, as high fees plus the price pressure that Fidelity generated against itself by herding with the Wall Street crowd took their toll. By contrast, investors in Vanguard received net annual returns of S&P - 0.6%. The gap compounds: Over 35 years Vanguard investors double their relative wealth. This gap drove John Bogle insane. But it did persist." What's wrong with Bogle's analysis? I'm not really well versed in the financial language so bear with me.

[derosnec]

i'll comment on your take on Myth #1 for now (because i need get back to cramming for a final)

your argument relies on showing that a few famous people have beat the market. that's not convincing.

then you cite Graham-Dodd's approach, which has been exposed by some heavyweights in financial research for its shortcomings.

this is a complex topic. you can't expect to make it in a few paragraphs. a discussion of the Modigliani-Miller theorem would take up several pages of posts alone. and that's just one tiny part of the whole thing.

as you know, academics have been debating this topic for decades. there are hundreds, probably thousands, of journal articles and papers with more numbers and evidence than we can possibly digest on this issue.

not trying to sound defeatist, bu this is an enormous topic with a mountain of data that needs to be analyzed (and a mountain of analysis that has analyzed that mountain of data).

[ConstantineX]

Let me see if I can grasp your argument in a primitive, layman form. Please feel free to correct, remand or revise. Your first myth states that the strong form of the Efficient Market Hypothesis is false. Irrational traders, coupling of sectors, movements of the market, bad calculation can generate alpha by these talented managers exploiting these idiosyncracies (Simons does this, I presume). Moreover, the length of time these traders have beat the market would suggest these people are finding or extracting real value in their strategies.

Beta is a measure of the "riskiness" of a particular portfolio's asset allocation. You claim that it is a myth that riskiness and earnings are even correlated. A comment from de Long's blog states "Most hedge funds are supposed to be low-beta. To the extent that their risk is truly idiosyncratic, it’s worth paying 2-and-20 even for lackluster returns, because they don’t add any risk to a portfolio", so one cannot even easily quantify the risk a particular portfolio has, and it isn't necessarily greater than the entire market's. Myth 4 states that even with greater "beta" from more exposure to different sectors of the market, one can arbitrage or hedge this risk and create value for the investor.

Appreciate your thoughts.

[ahnuld]

Nice post Naj,

I only read myth one, ill read the rest later. But I want to say that my comments in the EL D forum are in line with what you are saying here. I just think very very individuals can beat the market, but some definitely do and it is possible to beat it. I just think that out of the thousands of funds, perhaps only a few hundred are being managed by people who can.

[Stepping Stone]

The attempt to debunk Myth #2 was amusing. Try again.

(indexer)

[ConstantineX]

Looooove no substance posts.

[edtost]

thanks for this, nd. glad to see there are people bridging academic knowledge with actual experience.

[NajdorfDefense]

[ QUOTE ]
your argument relies on showing that a few famous people

[/ QUOTE ]

It absolutely does not, that is a very poor misreading of what I wrote, they are simply well-known examples of alpha generators who might be familiar to the layman. You should work on your reading comprehension, but I understand if you were rushed.

dero, saying 'there are two schools of thought' is a cop-out. Ibbotson and Chen did the definitive studies of hedge funds, using tens of thousands of data points, Positive alpha generation is predictive:
Ibbotson and Chen's academic paper on the topic of Hedge Funds--
http://papers.ssrn.com/sol3/papers.c...ract_id=733264

using Ibbotson's data for equity L/S funds only [Jan 1995- April 2006:
'L/S Hedge Funds create a <b>net alpha of 5.41 per year</b> [- that is net -], and overall return of 13.10%.'

All results are for dead + live funds, and do NOT include backfill data/bias. Alpha per HF strategy:
'Equity Neutral +1.94%
Fixed Inc Arb +3.91%
L/S Equity +5.41%.'

To avoid anyone's ingrained equity bias on this topic, simply use the Beta of 'the market', and also earn the 3.91% net alpha available in FI Arb funds. If you think loss aversion and groupthink is bad in equity-land, you should check out the bond market.

His most recent studies, using millions of daily/monthly datapoints on mutual funds over the decades, is similarly definitive and unrefuted. Given his 5 CFA Institute awards for excellence in research, awarded annually, plus his expertise in this field in both academic and real-world applications for 3 decades, you can't simply hand-wave him away. [and to attempt to without reading all his research on this topic is patently absurd]


The facts destroy any # of academic theories about 'how no one can beat the market.' I would agree most people could not beat the market after fees, but some clearly have, by a statistically significant margin. I get the feeling you are ignorant of the size and consistency of outperformance of those managers in the top 1% of 1%, decade after decade -- you would need 80 Quadrillion professional managers on Earth to explain away just 1 of them.

Regardless, either you are saying:
1) Alpha does not exist anywhere, which is patently false, or
2) Alpha is unobtainable by anyone.

If you really believed #2, you would buy your market beta for 2-3bps from BGI and short-sell listed hedge funds: RAB Capital, Soros, Man, Polygon, et al in London, there are hundreds more listed in Luxembourg.

In a 10% 'market' year, you'll make your 9.95% on beta, and also make another 3% by having shorted the '1 and 20' guys who will only be returning 7% in those years.
Or, run market-neutral, make Libor +3% over time. People would flood you with more capital than you could every possibly imagine.

[NajdorfDefense]

[ QUOTE ]
Nice post Naj,
but some definitely do and it is possible to beat it. I just think that out of the thousands of funds, perhaps only a few hundred are being managed by people who can.

[/ QUOTE ]

I wholeheartedly agree.

We're invested with a manager who's beaten the SPX by 40bps a year after fees for 15 years...with 65-75% less volatility.

Think about that...that should be impossible! Better market performance and alpha generation with a vol 3-4x less and no correlation -- I can't even think of a metaphor describing how insanely difficult that is, but it obliterates the notion of 'luck.'
Your finance professors [90% of them] will tell you it's flatly impossible. They are wrong.

However, I'm not here to debate the existence of managers I've been invested with for years, any more than I would debate a flat-earther or the Earth being the center of the solar system. John Neff's record of alpha generation alone over 384 months, crushing the index by a statistically significant margin speaks for itself.

One cannot maintain that '90% of managers trail the market by 2%' and also assume away people who net 50% a year for 2+ decades after trading costs. The data contradicts the theories quite easily obviously. It's even more obvious in the world of fixed income.