Last week we received a great response to our friend Ken Grant’s guest post “The 10 Commandments of Risk Management“. This week Ken builds on last week’s themes and introduces the concept of “Alpha” generation in ones portfolio. Ken uses AAPL’s 2012’s parabolic rise to all time highs to its near round trip to 52 week lows just this month to demonstrate the concept of positive and negative Alpha, and the implications on single position and portfolio risk management. The professor is in full effect here, so dig in and enjoy on an evening where it appears the financial world is fixated on tiny island with the population slightly less than that of Hawaii.
-Dan
[hr]
Redefining Alpha for Current Day Investment:
What Was Stock Performance and Selection is Now Management of Risk Premium
In my experience, few, if indeed, any, terms in the financial lexicon are as abused, confused and misused as Alpha. Over the years, its meaning has morphed multiple times, and, to add to the complexity of the chaos, each time, the legacy definition(s) with some justification, retain their original meaning and relevance.
In this article, we will explore these various usages and applications, as they have evolved, in my view on a rational path – towards a measure of efficacy of risk management for active portfolios.
The term Alpha first entered, nay, assaulted the investment universe as part of the Capital Asset Pricing Model (CAPM), introduced by the long-forgotten Jack Treynor in 1961, and subsequently co-opted by ubiquitous Nobel Laureate William Sharpe. However, as a proud graduate of the University of Chicago Graduate School of Business, I first encountered it through indirect counter-arguments socialized by, the fabulous non-Laureate Gene Fama (this is a travesty, but I try to remember that the Swedes in question actually gave the Peace Award to Obama, and even more scandalously, the Econ Prize to the increasingly unhinged and always wrong Paul Krugman a couple of years back), author of the academically critical, but (to be fair) problematic in the real-world Efficient Markets Hypothesis, which basically posits that market prices reflect all available information accurately at all times, and as such, either denies the existence of Alpha, or is fixes it for all time at a value of zero. In my mind, both CAPM and the Efficient Markets Hypothesis are bookends, in between which contain the real world of investment and money management — the dusty volumes that chronicle human behavior.
CAPM, in its original form, sought to separate an individual stock’s returns into two basic elements: 1) the performance of the market; and 2) the idiosyncratic contribution of the company itself to its equity valuation. Simply stated, it uses regression analysis to define an equity security’s market related performance, captured in the widely used (and almost as wildly abused) Beta statistic, and assigns the lion’s share of associated deviation there-from, the term Alpha. If you’re a regression-head like me (and God help you if you are) Beta is the slope of the best fitting line between market and individual stock performance, while Alpha is its intercept. Stocks that outperformed the product of their Beta and period to date market performance are deemed to have positive Alpha; those that underperformed this benchmark (called expected return) are assigned a negative ά.
To pick a high profile example, let’s return to those heady days of September 2012, when Apple sported a valuation of $700/share. Its price at the beginning of the year was around 405, so by the time it reached its peak (and before careening down in the following 2 subsequent quarters), it had generated a return of approximately 73%, against an overall market gain of roughly 16.5%. Further, a review of historical data indicates that its Beta at the time was approximately 0.815. Applying CAPM, we find that this iconic company had an expected return of around 13.5% (16.5% * 0.815) and that its excess return, or Alpha, was over 60% — an astonishingly large number. CAPM would tell us that this was substantially due to the outsized success of its mobile product lines, and an expectation among market participants that such differentiation would carry forward into the future.
But, as Dr. Fama: Robert R. McCormick Distinguished Professor of Finance at the University of Chicago’s Booth School of Management, would no doubt be quick to note, Apple’s Alpha-driven, Newtonian tumble back to earth since it hit the 7 handle, was an Alpha train wreck. At its low point – reached just days before I typed these words, it hit a level of 420 – an alarming 40% drop in less than 6 months – particularly for what had been arguably the world’s favorite stock for the past several years. Moreover, this decline transpired during a period when the benchmark SPX index actually rose, in steady, if un-dramatic fashion – from 1465 to 1525 (4.2%). Over this interval, AAPL’s expected return, given that it still supports a Beta of approximately .75 (we’ll discuss problems with the Beta calculation in a later installment), in round numbers, was a little over 3%, and had it done nothing more than match its Beta-adjusted market performance, it would be trading 300 points higher than its current price projects. Thus, its Alpha from Mid-September, 2012 through early March was -43% — again in round numbers.
Finally, if we consider the entire timeframe from January 1, 2012, through March 4, 2013 (i.e. the date it closed at its 420 low), the overall market was up over 21%, AAPL’s expected return (given an average Beta of .80) would be slightly over 16%, but instead, the stock was up a mere 3.7%, and thus has generated over 12% of negative Alpha for the period from January, 2012 till the point of its 420 print..
Somewhere, presumably in the Silicon Valley section of heaven, Mr. Jobs must be heaving a forlorn sigh.
As he cries his heavenly tears for recent “Apphla” trends (a term I just coined), we the living can focus on more earth-bound lessons, which, if they don’t fascinate you, certainly intrigue me. Specifically, the story of AAPL for the last 5 quarters has been a round trip from the low 400s, to 700, and back, with stalwart investors booking a piddling 15 point gain for all their troubles. The ride up produced a 60%+ ά, while the backslide generated a negative Alpha of 43%. Yet, somewhat counter-intuitively, when I combine a 9-month 60+ Alpha climb with a 5-month ά reversal of lesser magnitude, I get double digit negative ά across the entire interval. To put this in further perspective, if we assume a continued (albeit flatter) upward trajectory for the markets across all of 2013, and that by year-end, it closes with an incremental net gain of an additional, say, 6% (taking the full-year rally to 15%), along with a sustained Apple Beta of .80, then the stock’s expected return for the remainder of the year would be just under 5%, and, given its 5 quarter ά result of -12%, AAPL would actually require a remainder of the year rally of 17% — just meet its expected return, zero Alpha threshold for the 2012/2013 interval as a whole. In price terms, under the assumptions listed above, the stock would need to approach 500 simply to reach the ά Mendoza Line of 0 for the two years in question. And that is assuming the market gains only 2/3rds in the remaining 9 months of what it achieved in the 1st Quarter.
All of the above speaks to the asymmetry of risks across a P/L cycle, in particular, the deeply debilitating impacts of holding static portfolios through periods of stock-specific or market-wide impairment.
There are important messages we can take from this Newtonian morality tale, as they apply to the world of modern investment: one much different than that in which JT/WS created CAPM, and which GF rebutted several years later.
The evolution, as it applies to Alpha, has occurred in phases. First, as investors migrated towards managed, diversified portfolios, long-only portfolios, such as those embodied in mutual funds, Betas evolved into one reflecting a mix of securities, and any Alpha – positive or negative – was a reflection of the stock selection acumen of the portfolio manager in question. However, since most of these “diversified” portfolios operate in a relative return world, and deviate only nominally from their benchmarks, Alpha contributions, both positive and negative, diminish to microscopic proportions. A benchmarked manager will typically generate ά in the +/- 2% range at most – a far cry from that of a talented and fortunate market operative who, with proper foresight, might’ve ridden Apple on the long side from 400 to 700, turned around and shorted it, and copped over 100% Alpha in a single name – all in a 5 quarter interval.
But markets keep evolving, most applicably from a benchmarked, relative return construct, into an absolute return paradigm. What’s different about this world? Well, among other things:
- Financial information and analysis are more quickly disseminated, and this has marginalized stock selection, as viewed in isolation, as a key differentiating factor of portfolio performance.
- Many capital pools have both long and short components to their portfolios.
- On the whole, there are much more active portfolio adjustments on the margin (e.g. the sizing up and down of postions) than in eras past.
- The risk appetite of global investors is changing at a more rapid pace than ever before – mostly due to the uncertainties and rapid evolution of macro events, and as such, at varying points during an investment cycle, the price extracted for risk transference from one economic agent to another rises and falls accordingly, in difficult to discern magnitudes.
For the purposes of this discussion, we will define the price of risk transference as the “risk premium”, and it is indeed a wily and unpredictable beast. It expands and contracts virtually at its own whim, and typically without advance notice. Moreover, when it rises, it brings with it a number of unpleasant impacts:
- Liquidity diminishes, and in some instances, disappears altogether.
- Prices change in ways that decouple from varying opinions regarding “true” valuation.
- Correlations across certain instruments rise, while falling across other market segments.
I think the last point is worth discussing in greater detail. Many market participants equate high risk premium markets with across-the-board increases in the correlation of pricing across all tradable instruments. This, in my experience is a dangerous fallacy. If ALL instruments moved in lockstep during periods of risk turbulence, then, among other things, hedges would work better than they historically have performed. Those who have been actively invested in the markets during these periods know this from their experience, so the next time someone tells you on a bad tape that it’s high correlations that are killing them, I suggest you give them a cryptic, somewhat disgusted look, and walk away, lest you say something impolitic.
To explain how all of this has impacted the Alpha paradigm, an example best suits our purposes. So, for our objectives, let’s assume that the net Betas for most portfolios run somewhere around 50% of their investment levels, and use this figure because it is easy for the mathematically challenged (including your humble correspondent, who’s once fine quantitative chops are eroding).
Using these parameters, and applying it to a roughly 8% year-to-date gain in the SPX, our 50% β portfolio would generate an expected return of 4%; anything above that would be indicative of positive Alpha and visa versa.
But these one-way-ticket-up markets, while lucrative for most, are not particularly useful for our goal of erudition, so let’s envision, hard though it might be, a market that experiences a 10% correction. Our 50% Beta portfolio would have an expected return of -5%, and if it beat this number, even if it lost money in the process, it generated positive Alpha for the period.
But here’s the rub: in my experience, it is more difficult to produce Alpha outperformance on a strong/steady tape than it is on during a cycle that is less stable. This may seem counter-intuitive, but it’s true. In a rally such as that which has characterized the last 5 quarters – a period which has produced approximately 19% returns, with a maximum drawdown of roughly 5%, creating a portfolio that outperformed its expected return required some doing. There are those out there that achieved this, but they had to fight for every inch of ά they achieved – mostly by picking outperforming stocks on a tape where virtually all names were rising.
However, in a cycle that begins with “down” market which then reverts to relative stability (as is inevitable), producing positive Alpha (or, at any rate, better Alpha performance than your peers) becomes a mere matter of reducing exposures when the risk premium experienced a discrete jump (there were other ways, but trust me, this is the easiest one), and using this advantage to scoop up mispriced securities when normalcy begins to set in. The most recent example which comes to mind is the 17% drop from mid-July until late August. Not only did most investors lose money, but most gave back significantly more than their expected return.
In turn, the reason for this is that when Congress, after passing 50 debt ceiling increases in as many years without so much as earning a single mention in the press, turns the 51st into a complete fiasco, and one of the world’s leading rating agencies makes good on a threat that all God’s children assumed was a bluff, by downgrading U.S. Sovereign Debt – all in the space of 2 weeks, investors of every stripe run for the hills. Those S&P puts you thought you were so smart to use as a hedge move up in value, but it only modestly offsets the beating you took when the stocks that you owned, with finite liquidity and idiosyncratic risks, saw their valuations sink like a stone. If you didn’t act quickly and proactively, you might’ve easily encountered double digit draw-downs in the blink of an eye. However, if you recognized the rising risk premium – by no other means than tracking your own Alpha performance on a daily basis, and beating your peers in the race to a lower risk profile, then you did well – both for yourselves and your investors. In our example, by the time the markets reached their “clearing price” of 1100 on the SPX (early October, 2011), the disease of negative Alpha, in our observation, forced many competent market participants, who did not, per se, manage their exposures with maximum effectiveness, to sell at these lows.
In the subsequent 6 quarters, the index rallied 450 points (41%) with a maximum peak-to-trough correction of less than 10%. Those who outperformed their peers during the –ά 3rd Q3 11 correction, from a risk management perspective, almost to a one, captured more of the upside (which featured many oversold names that the wise and prudent out there were able to purchase from forced sellers)than did those who operated with sub-optimal risk effectiveness. To reemphasize, it was not necessary to generate positive returns during this cycle, or even, for that matter positive Alpha; rather, mere Alpha outperformance relative to peers placed risk takers in a position to outperform in subsequent cycles, often by a wide margin.
“Risk Off” periods are inevitable, and there is both bad and good news embedded in this reality. The bad news, of course, is that they are unpleasant to endure, and only the fortunate and/or clairvoyant generate positive returns during these cycles. The good news that because: a) they don’t last forever; and b) they leave many pricing inefficiencies in their wake, they set the stage for execution of many of the best themes available across an investment sequence. Identifying the mis-pricings when the waters becalm themselves is child’s play; avoiding getting carried out during the preceding rip-tides is, in my experience, as big a differentiator of raw and risk adjusted performance as any of which I am aware in the current, highly unstable universe of risk assumption and management.
So, to summarize, Alpha is an unstable beast, and one that at this point, as well as going forward, is much more a reflection of the sizing and timing elements of exposure management, against a highly unstable risk environment, than it is a measure of position selection wizardry.
Stated more simply; for individual portfolios, positive Alpha implies sound risk management; negative Alpha indicates the opposite paradigm prevails. For the market in general, widespread negative Alpha is an indicator of rising risk premium, while broad-based positive Alpha is elusive, and, most importantly, most easily achieved by managing your risk through difficult period better than other market participants during the tough times, thereby placing yourself in the best position to capitalize on myriad price inefficiencies that emerge during the risk reduction cycle. It is critical to note, however, that the worst time to try to capture pricing inefficiencies is during impaired market cycles; one must, as suggested above, wait till the waters calm, before casting their lines into a sea of hungry fish.
I feel I would be remiss if, prior to taking my leave, I didn’t offer some thoughts as to the best way to manage through the redefined Alpha construct, so let’s have a go at some managerial approaches, shall we?
Be Aware of the Risk Premium – As it Applies to Your Portfolio and the General Market. Whether you trade actively or manage a static portfolio, your risk profile is changing all the time. Even for active traders, tools are available to help you to partition the contribution of these changes – between shifting market sentiment and the adjustments you make to your holdings. One of these is the tracking of Alpha on a daily and cumulative basis. If you find, for instance, yourself entering a period where you are routinely underperforming your expected return, you are at liberty to hang your head and proclaim to yourself that you’ve lost your golden stock selection touch.
But more likely than not, this would be the wrong conclusion to draw. You’re probably as much of a market golden boy (or girl) as ever, but you are experiencing an interval where liquidation is occurring in a price insensitive manner. If you can detect these patterns, then you can make the appropriate adjustments that will allow for the capture of outperformance opportunities that are likely to emerge when the fire-sale liquidation cycle ends as it inevitably will.
Adjust Your Exposures Downward During Negative Alpha Cycles. A blinding glimpse of the obvious perhaps, but less well-executed than understood as a concept, in my experience.
When Forced to Reduce Your Exposures, Seek to Optimize the Lower Risk Profile. In periods of broad-based negative Alpha, your favorite themes may appear to be the best place to perform the risk reduction surgery. This is another way of expressing the (in my judgment) highly dubious risk management concept of selling your losers while letting your winners run. This may be a good strategy on a good tape, but it’s a bad one on a bad one. If the market is falling apart, names and themes that you already believed to be mispriced are likely to further skew themselves from rational valuation levels.
As such, across all market cycles, when risk reduction is called for, I continue to recommend that it be using what I like to refer to as the Vince Lombardi “Gentlemen, this is a football” strategy. It involves going through your names, ranking what you want to hold – long or short on a “conviction” basis – at current valuations, and irrespective of your period to date p/l in these names. The strategy does not, per se, imply that you will liquidate your losers and hold on to your winners; however, my strong empirical experience is that investors’ beliefs on valuations do not (or at any rate, should not) be deeply impacted by changing risk conditions. As such, there’s probably more “juice” in your undervalued longs and your overvalued shorts than there is in the portions of your portfolio that are “working”, and, in turn, this implies that you will typically better off getting out of winners and holding on to losers.
However, before you undertake any of these activities, it behooves you to eliminate the superfluous portions of your portfolio. I am constantly amazed, across the full spectrum of investment pools, the degree of content consumed by positions where those responsible have no reason, or can’t remember why. they put them on in the first place. Start your risk reduction with these, and then move to “core” positions, reducing exposure to them in ascending order of their upside potential.
When in Doubt, Gross Down. The simplest, and typically most effective, means of reducing risk during negative Alpha cycles is by purely reducing the size of the investment portfolio. If you’re deeply confused by it all, just cut everything in equal proportion. As a matter of embedded arithmetic, the volatility you will experience on a forward-looking basis will be reduced by at least the same percentage (more if you achieve an improved liquidity profile through the reduction process).
Don’t Expect Your Hedges to Help You Much. Many investors take great comfort in the puts or short ETF positions they hold, but they seldom work in a negative Alpha environment, due, in part to the above-mentioned “correlation” fallacy. Out of the money puts, in particular, tend only to pay off when market conditions deteriorate in material and abrupt fashion, while ETF-type hedges are likely, on an impaired tape, to provide only partial offset to the bleeding that is taking place in the flower of your portfolio.
One final point on hedges: price matters. Put them on when their cheap; not when the world is falling apart and everyone is scrambling for offsets. Sadly, most investors do the exact opposite, and if I had a penny for every million dollars that was lost due to this inefficiency, well, we won’t get into that.
Use Options Markets and Do So With Nuance. In a negative Alpha environment, volatilities rise, and do so unevenly, so there are typically myriad ways to exploit options markets, to your incremental benefit, across these cycles. When vol is high, selling calls against your long positions is a tragically underutilized tool in portfolio management. Similarly, if you are nervous about your Alpha, you can often swap a portion of your direct exposure into options, capturing the majority of upside while limiting downside.
In addition, particularly in a price insensitive, negative Alpha environment, there are typically myriad ways to construct combined options positions, at “below market “ costs, with disproportionate upside benefits. This paradigm is both reflective of, an enabled by, the deeply inefficient manner in which options are traded by even the most sophisticated of investors.
I’m running out of steam here, ladies and gentlemen, and simply conclude by encouraging you to think of Alpha as both a risk barometer, and a tool for capital preservation that, if understood, can place you in a position to capture price inefficiencies at optimal points in an investment cycle. Stop thinking about it as a measure of your genius as a stock picker, and start using it as a roadmap of current and future risk conditions.
If you do so, I suspect you’ll come out better off than if you don’t, and perhaps most important of all, you’ll have earned my respect, which is something, perhaps counter-intuitively, that someday, some way, you’ll be able to take to the bank.
Others have; and you can too.
Peace Out.