Long Cheap; Short Expensive. Buyer Beware.

/Long Cheap; Short Expensive. Buyer Beware.

Long Cheap; Short Expensive. Buyer Beware.

By | 2017-08-18T17:01:24+00:00 October 31st, 2014|Research Insights|19 Comments

A recent New York Times article examines the story of Joel Greenblatt, author of “The Little Books That Beats the Market.” Joel has leveraged this book and is now managing money using a similar strategy for Gotham Asset Management. Joel is, of course, an investment legend at this point. We have been fans of Joel because he follows the goals of our firm, which is to empower investors through education. All that said, at Alpha Architect there ARE NO SACRED COWS.


Some of his newer mutual funds involve a hedge-fund investment strategy that goes long cheap stocks and short expensive stocks.

Here is the quote from the article:

At a meeting on Oct. 1, with 10 brokers in a cramped conference room in his Madison Avenue office and 40 more listeners on the phone, a tanned and confident Mr. Greenblatt told them, “We buy the cheapest we can find and short the most expensive.” With copies of his book stacked behind him, he said the general methodology had been “back-tested back to 1990” and “has worked for 30 to 40 years, no question.”

While not taking a beta-neutral bet, the largest fund (Gotham Absolute Return Fund) involves having a long exposure of 120%, and a short exposure of 60%.

So how has this fund done?

Here are the results:

2014-10-30 14_03_37-Fund Offerings _ Gotham

By any estimation, on a market-exposure adjusted basis this fund has done a great job since its inception.

In expectation, a fund with a 60% net long exposure would expect to return 60% times the market return plus the risk free rate. During this time period (with the market return = 20.15% and risk-free rate essentially equal to 0%), the expected return would have been (60%)(20.15%) + 0% = 12.09%. Gotham Absolute’s 18.85% has after-fee out-performance of over 6.75% on an annualized basis–not bad!

So even after the large management fee (2.00%) in the prospectus, and a 1% redemption fee (within 90 days), is this a free lunch?

Investigating long cheap, short expensive strategies

We decided to do a simple analysis of Mr. Greenblatt’s “magic formula.”

  • Variable 1: EBIT/TEV (Total enterprise value)
  • Variable 2: EBIT/Total capital

To avoid smaller firms, we eliminate all firms below the NYSE 40th percentile, which was around $1.95 billion on 12/31/2013. On 12/31/13, this leaves a universe of around 1050 firms.

All returns are total returns and include the reinvestment of distributions (e.g., dividends).

We then rank the firms on the two variables, and form quintiles of the average rank. The quintile portfolios are formed on 6/30 each year, held for a year, and equal-weighted.

To replicate the Gotham Absolute Return Fund, we buy the top quintile (~200 firms) with 120% exposure, go short the bottom quintile (60% exposure), assume a management fee of 2.00%, and a short rebate of 0.25%.

Here is how our replicated strategy compares to the live returns (from 9/1/2012-12/31/2013).

2014-10-30 15_33_34-Microsoft Excel - MF long short analysis v01.xlsm

The results are hypothetical results and are NOT an indicator of future results and do NOT represent returns that any investor actually attained. Indexes are unmanaged, do not reflect management or trading fees, and one cannot invest directly in an index. Additional information regarding the construction of these results is available upon request.

While the replicated strategy returns are not exactly the same, they are clearly very similar, with nearly identical compound annual growth rates (CAGRs) and a 84.46% correlation over the 16 month period. For such a short time period, this level of similarity suggests a pretty close relationship between the simple magic formula and the version that Gotham claims to be implementing, which is “enhanced” by a team of analysts who make adjustments to account for esoteric valuation considerations such as pension obligations and legal claims.

So what about Greenblatt’s claim this strategy has worked for “30 to 40 years, no question.”

What Happens to Long Cheap, Short Expensive, Historically?

Here are the returns from 1/1/1960-12/31/2013, using the same fee assumptions outlined above:

2014-10-30 15_39_28-Microsoft Excel - MF long short analysis v01.xlsm

The results are hypothetical results and are NOT an indicator of future results and do NOT represent returns that any investor actually attained. Indexes are unmanaged, do not reflect management or trading fees, and one cannot invest directly in an index. Additional information regarding the construction of these results is available upon request.

As Mr. Greenblatt would say, “gee, that’s interesting.” Long cheap, short expensive works, but you can expect to nearly go bust along the way. As I told my cable guy recently, “that’s not the deal that I signed up for.” So much for the seemingly sophisticated hedging techniques and “hedged” risk. What’s the point of a big short book if it doesn’t protect you from the big drawdowns? A -52.01% drawdown isn’t low-risk on any planet I’m familiar with, and in fact this strategy has an even worse drawdown than long-only market exposure. Wes and I outline just how risky many “anomaly” strategies can be in a long/short context.

The large maximum drawdown occurs during the internet bubble (5/1/1998 – 2/29/2000), which is a bad period to be long value stocks and short growth stocks.

Will there ever be another internet bubble? Who knows? But investors should be aware that if they face similar market conditions, they can expect to lose over half of their money in the long/short strategy proposed by Gotham, or in investments with any manager peddling long cheap, short expensive type strategies.

Is there a way to improve the Long Cheap/Short Expensive strategy?

One suggestion would be to go the cheaper route, and form a DIY portfolio, as opposed to paying 2% management fee.

Another option is to simplify the approach and build the portfolio by going long the top quintile of cheap names (120% exposure) and shorting the SP500 (60% exposure). We assume a 0.25% short rebate and a 0.20% annual transaction cost, as we only need to rebalance the long portfolio annually, and shorting with a SP500 future has minimal costs.

Here are the returns from 1/1/1960-12/31/2013:

2014-10-30 16_09_17-Microsoft Excel - MF long short analysis v01.xlsm

The results are hypothetical results and are NOT an indicator of future results and do NOT represent returns that any investor actually attained. Indexes are unmanaged, do not reflect management or trading fees, and one cannot invest directly in an index. Additional information regarding the construction of these results is available upon request.


The simple DIY strategy is better on any performance metric. The DIY approach (first column) has a higher CAGR, Sharpe and Sortino ratios, and a lower maximum drawdown.

Overall, the Magic Formula Long/Short strategy is not bad by any measure, but is by no means a free lunch. In time periods such as the internet bubble, losing half of your money seems quite possible.

The fees (2%) are extremely high, and a simple DIY strategy would seem to give better risk-adjusted returns and has the benefit of being relatively easy to implement.

If additional fees (such as fees from brokerage houses) are added on top of the 2% management fee, the DIY option may be worth the headache of implementing such a strategy.

Last (and maybe most important), the mutual fund setup of Gotham means taxes will be distributed each year, so after-tax returns will, in general, be lower.



  • The views and opinions expressed herein are those of the author and do not necessarily reflect the views of Alpha Architect, its affiliates or its employees. Our full disclosures are available here. Definitions of common statistics used in our analysis are available here (towards the bottom).
  • Join thousands of other readers and subscribe to our blog.
  • This site provides NO information on our value ETFs or our momentum ETFs. Please refer to this site.

Print Friendly, PDF & Email

About the Author:

Jack Vogel
Jack Vogel, Ph.D., conducts research in empirical asset pricing and behavioral finance, and is a co-author of DIY FINANCIAL ADVISOR: A Simple Solution to Build and Protect Your Wealth. His dissertation investigates how behavioral biases affect the value anomaly. His academic background includes experience as an instructor and research assistant at Drexel University in both the Finance and Mathematics departments, as well as a Finance instructor at Villanova University. Dr. Vogel is currently a Managing Member of Alpha Architect, LLC, an SEC-Registered Investment Advisor, where he heads the research department and serves as the Chief Financial Officer. He has a PhD in Finance and a MS in Mathematics from Drexel University, and graduated summa cum laude with a BS in Mathematics and Education from The University of Scranton.
  • IlyaKipnis

    This is terrific analysis. Cheap can get cheaper, and expensive can get more so. This is usually not the case, but if you bet the farm on this, one particularly bad event can wipe you out. In this case, if you even leveraged the little book’s strategy 2:1, you would have been wiped out in the tech bubble.

  • yup. Not sure many folks understand just how risky long/short can be…

  • Antacular

    Posts like these is why this is the only investment blog I read regularly. Simply excellent.

  • sb

    You don’t have to put all your money in this one strategy. Authors assume that Greenblatt’s fund simply is the Magic Formula, which I believe he has said the fund is not. They have a much more involved process. In the past he has said you don’t simply short the worst stocks using the MF because you would go broke. Not sure this shows you would be wiped out as I see the worst drawdown being 52%. I’d say pretty impressive performance since we have a lot of stocks with similar characteristics now as in the late 90s tech bubble with the current Internet bubble 2.0.

  • Agree completely. And hedged strategies can work well in a broader portfolio context. Great point. We are simply highlighting the risk/return associated with these stand-alone strategies. As a portfolio investment, a lot of their “risk” might be washed away via diversification.

    The analysis we conducted does not suggest that their process is much different than the simple magic formula. The data argue they are statistically the same. My guess is that one needs to “razzle dazzle” things a bit to justify higher fees (ie. call it proprietary and conduct a lot of activity).

    In a long/short context, drawdowns are more important than in a long-only non-margined context. Remember, when you are long/short you are typically long fully paid names and you use those long names as your collateral for the short book. In other words, you are using leverage. A 50%+ drawdown in a L/S context usually means margin calls and forced liquidations. In a long-only context it simply means you want to jump off a bridge, but at least your banker can’t liquidate your positions…

  • sb

    Thanks for the reply, Wesley. I was thinking the drawdown meant that the investor lost 52% of his invested capital in the fund. But if I’m understanding you it really means something like 52% of the 120% long and 60% short. Which gets to 90% or so lost?

  • Hi SB,

    It means the “fund” lost 52%. Sorry for the confusion.

    Main point is that long/short–or perturbations thereof–are leveraged trades, whereas long-only trades are not leveraged trades.

    At those maxdd levels, and with the right conditions (e.g., a flat long book and a short book that is 50% against you), there is a decent shot the fund manager would get a margin call because the broker would require additional collateral to maintain the short positions.

    We talk about the issue with large drawdowns in the context of long/short strategies in our paper:


    One of the key findings (see table 5) is that L/S systems usually have the best returns following a massive drawdown event in the L/S strategy (ie. after everyone in the trade is bankrupt).

    Of course, nobody can actually earn the “bounce back” returns because everyone in the trade has usually busted out via a margin call. In order to exploit the returns on the L/S trade you’d need to have a broker with an amazing temperament for risk and an ability to maneuver around regulatory mandates on collateral requirements.

  • sb


  • Nathan

    Thanks for this very informative post. Would you know what the drawdown would have been for 175% long and 75% short (i.e. GENIX, Gotham Enhanced Return Institutional GENIX)?

    When the original Formula Investing Focused US fund was shut down all shareholders (including a friend of mine) were automatically transferred into GENIX.


    P.S. According to Morningstar, these new Gotham mutual funds exclude Financials, Utilities, Energy and Materials. Did your backtest exclude those sectors?

  • Nathan,

    I don’t have it off hand, but my guess is the drawdown for a 175/75 would be 50-55%. Jack can probably respond with details. He worked the data on this one…

    In this particular backtest we only exclude financials. That said, we’ve done tests on just about everything you can imagine when it comes to value-based quant strategies. With the magic formula, various perturbations all tell the same general story plus/minus noise.

  • Nathan

    Thanks Wes. How would I go about contacting Jack? The email link by his profile doesn’t seem to work.

  • He follows the comment streams. Jack is going to do a follow on post with a variety of perturbations.

  • Jake


    Have you run any analysis where you buy the cheapest decile (or better yet the cheapest 5%) based on EV to EBITDA or EBIT subject to an ROIC cutoff value (e.g. low EV/EBITDA with ROIC’s greater than X% – perhaps 20%?). This way you would weight the portfolio by value (rather than a 50-50 combination like in the magic formula) and add a quality metric. This would test whether 1) the ROIC metric adds value and 2) the highest returns are actually from buying bad companies at cheap prices that then turn around rather than good companies at cheap prices.

    Love your site!

  • Jack Vogel, PhD

    Thanks Jake!

    We have run these tests, and plan on posting sometime in the near future. Good suggestion!

  • Martin Bligh

    Do you have a way to run the same numbers for their market neutral fund (GONIX) ?

  • Jack Vogel, PhD
  • Martin Bligh

    Ouch. Great analysis, thanks.

  • Adam Kearny

    Very thoughtful and rigorous work. The only problem is that it’s a bit too “30,000 feet” in that you presume it’s mostly about EV/EBITDA (in isolation) because that’s matched up over the past two years. The Gotham Funds expressly use sector concentration caps and position caps that would have avoided most of the massive draw-down you imply. The 1998-99 bubble was concentrated in tech and to a lesser extent large-cap growth. The worst that would have happened is that Gotham Funds would have scaled into tech shorts until they grew to about 5-10% net short (in the tech sector) and would then have been stopped out of them, incurring a 5-10% permanent capital impairment in the tech sector but no further losses in that sector. Therefore, your back-tested example seems unrealistic (though it is certainly the pertinent historical example to consider).
    Also, from a more common sense perspective, these guys aren’t dummies–Greenblatt was managing a hedge fund through the late 1990s and is obviously well aware of the potential risk; I have a hard time believing that this current strategy wasn’t stress-tested for such a period (though admittedly, it’s theoretically possible we could get a mega-bubble in junky garbage across all sectors that puts 1999 to shame). In studying their Oct. 31st holdings, they’ve been short a bunch of small-cap biotechs the past four months most of which have soared since Oct. 31st (some have doubled) as biotech is in a parabolic bubble (XBI returned 42% last year). Nonetheless, in estimating the net losses to GONIX to-date from biotech/pharma since Oct. 31st report, I’m coming out with only about 30 basis points.
    One thing that jumps out at me from their Oct. 31st portfolios is that most of the shorts are small-to-mid caps. This is in-line with Greenblatt’s public comments over the past year or so about valuation dispersion between small-caps and large caps (further backed by GMO/Jeremy Grantham, Leuthold Group, etc.). You also notice that Gotham Funds tends to do worse on days where small-caps strongly outperform large-caps. The biggest systematic risk to Gotham Funds right now is that small-caps go on an extended tear over the next year or two, leaving large-caps in the dust. Anything is possible, but I would note that in 1998, large-caps trounced small-caps by 3000 basis points while in 1999, they were very close. Small-caps typically outperform (excluding beta-adjustments) coming out of a bear market and underperform during the later stages of a bull market as financial conditions tighten (like right now–see credit spreads action; or even 1998-99).
    Your basic point about extreme conditions and drawdown risks is well-taken, but I’m not convinced by this over-simplified example. There’s no historical example I’m aware of involving small-caps outperforming large-caps by gargantuan amounts (particularly from current valuation levels) which is what it would take to create the type of drawdown you suggest. I would also note that the shorts have already been annihilated (Chanos was down double-digits in 2013, and the “buy the most-shorted” strategy went gangbusters in 2013, early 2014–per your research, this suggests that now might be a good time to be long-short. There’s also anecdotal evidence that long-short is out-of-favor–long-standing liquid alt long-short TFSMX has posted disappointing results over the past few years and accepts new money again after closing a few years ago following a wave of performance-chasing / popularity for weathering the 2008-09 crisis so well.

  • Adam,
    Appreciate your comments. We’ve done a lot of research in this space and are familiar with many of the points you mentioned. They are certainly thought-provoking. We’ll see how it all plays out over the next 10-20 years.