Walter: Did you learn nothing from my chemistry class?
Jesse: No. You flunked me, remember, you prick? Now let me tell you something else. This ain’t chemistry — this is art. Cooking is art. And the shit I cook is the bomb, so don’t be telling me…
Walter: The shit you cook is shit. I saw your setup. Ridiculous. You and I will not make garbage. We will produce a chemically pure and stable product that performs as advertised. No adulterants. No baby formula. No chili powder.
Jesse: No, no, chili P is my signature!
Walter: Not anymore.
— Breaking Bad, Season 1, Episode 1
“There was only one decline in church attendance, and that was in the late 1960s, when the Vatican said it was not a sin to miss Mass. They said Catholics could act like Protestants, and so they did.“
— Rodney Stark, Ph.D.
She should have died hereafter;
There would have been a time for such a word.
To-morrow, and to-morrow, and to-morrow,
Creeps in this petty pace from day to day
To the last syllable of recorded time,
And all our yesterdays have lighted fools
The way to dusty death. Out, out, brief candle!
Life’s but a walking shadow, a poor player
That struts and frets his hour upon the stage
And then is heard no more: it is a tale
Told by an idiot, full of sound and fury,
— William Shakespeare, Macbeth, Act 5, Scene 5
“I can’t do it if I think about it. I would fall down, especially if I’m wearing street shoes,” he said, laughing. “It wasn’t something I did because I wanted to. I didn’t even know I did that until someone showed me a video.”
— Fernando Valenzuela about his unique windup to the LA Times (2011)
Baseball was in the midst of a crisis in 1981.
In the years prior, competition for talent in larger markets had driven player salaries higher and higher. This caused owners to seek increasing restrictions on free agency. The players’ union went on strike in June, right in the middle of the season. Fans were furious, and mostly with the owners, as is the usual way of things. We still hate millionaires, of course, but we positively loathe billionaires. While the strike ended by the All-Star break in early August, work stoppages and disputes of this sort have often been the signposts of baseball’s long, slow march to obscurity against the rising juggernaut of American football and the sneaky, if uneven, popularity of basketball. It was not a riskless gamble for either party, and as future strikes taught us, the aftermath could have gone very badly.
But not this time. You see, baseball had a secret weapon to quickly bring fans back after the 1981 strike: a “short fat dark guy with a bad haircut.” His name was Fernando Valenzuela.
Fernando was an anomaly in another long, slow march — that of baseball’s transition from a pastime to something more clinical, more analytical, more athletic. We were at a midpoint in the shift from the everyman-made-myth that was Babe Ruth or the straight-from-the-storybook folk hero like Joe DiMaggio to the brilliant, polished finished products of baseball academies today. Only a few years after 1981, we would see the birth of the new generation of uberathletes in Bo Jackson, a man who many still consider among the most gifted natural athletes in history. Only a decade earlier, the top prospect in baseball was one Greg Luzinski. The two weighed about the same. Their body composition was just a little bit different.
Fernando was certainly a physical throwback of the Luzinski variety, but so much more. He was a little pudgy. His hair was, long, shaggy and unkempt. More to the point, everything he did was inefficient, out of line with trends in the league. His windup was long and tortured, with a high leg kick that reached shoulder level in his early years and chest level in his older, slightly chubbier years. It featured an unnecessary vertical jerk of his glove straight upward near the end, and most uniquely, a glance to the heavens that became a signature of Fernando-mania. To stretch the inefficiency to its natural limits, his most effective pitch was a filthy screwball, a pitch that had been popular for decades but had already significantly waned by the early 1980s. Fathers and coaches taught their sons that it would hurt their arms (which a properly thrown screwball does not do), and by the late 1990s the pitch that ran inside on same-handed batters was all but extinct, except in Japan, where a very similar pitch called the shuuto continued to find adherents.
There were many reasons he captured the national imagination. He was a gifted Mexican pitcher in Los Angeles, a city full of baseball-obsessed Mexican-Americans and migrant workers. He was also truly marvelous as a 20-year old rookie in 1981. His stretch of eight games between April 9th and May 14th still ranks as one of the most dominant in history. Eight wins. Eight complete games. Five shutouts. Sixty-eight strikeouts. And that was how he started his career!(1)
But more than anything, I think, it was the pageantry and the spectacle of it all. The chubby, mop-top everyman who came out of nowhere with a corny sense of humor, who threw from a windup out of a cartoon, who threw a pitch that nobody else threw anymore. It was inefficient and ornamental and just so unnecessary — and we loved it. I still do. It was even how I was taught to pitch growing up. My father told me and instructed me to throw with “reckless abandon”, and so in my windup I would rotate my hips and point my left toe at second base before kicking it in a 180-degree arc at a shoulder level, nearly falling to the ground from the violent shift in weight after every pitch.
Alas, the efficiency buffs who disdained such extravagances were and are mostly right. While Valenzuela had a long and decent career, the greatest pitchers of the modern era — Roger Clemens, Pedro Martinez and especially Greg Maddux — all thrived on efficient mechanics and a focus on a smaller number of high quality pitches.(2) While a screwball is nice, and in many ways unique, it also isn’t particularly effective as a strikeout pitch in comparison to pitches with more vertical movement like, say, curveballs, split-finger fastballs or change-ups, or pitches that can accommodate lateral movement AND velocity, like sliders and cut-fastballs.
There’s a lesson in this.
As humans, especially humans in an increasingly crowded world where we can be instantly connected to billions of other people, the urge to stand out, to carve out a different path, can be irresistible. This influences our behavior in a couple of ways. First, it drives us to cynicism. Think back on the #covfefe absurdity. If you’re active on social media, by the time you thought of a funny #covfefe joke, your feed was probably already filled with an equal number of posts that decided that the meme was over, using the opportunity to skewer the latecomers to the game. Those, too, were late to the real game, which had by that time transitioned to new ironic uses of the nonsense word. A clever idea that is shared by too many quickly becomes an idea worthy of derision. And so the equilibrium — or at least the dominant game theory strategy — is to be immediately critical of everything.
It also makes us inexorably prone to affectation. We must add our own signature, that thing that distinguishes us or our product; the figurative chili-powder-in-the-meth of whatever our form of productive output happens to be. Since we are all writers of one sort or another now, we feel this acutely in how we communicate. When part of what you want to be is authentic in your communication, our introspection becomes a very meta thing — we can talk ourselves into circles about whether we’re being authentic or trying inauthentically to appear authentic. But we’re always selling, and while our need for a unique message has exaggerated this tendency, at its core it clearly isn’t a novel impulse. People have been selling narratives forever. But if there’s a lesson in Epsilon Theory, surely it is that successful investors will be those who recognize, survive and maybe even capitalize on narrative-driven markets — not necessarily those whose success is only a function of their ability to push substance-less narratives of their own.
Perhaps most perniciously, our urge to stand out is also an urge to belong to a Tribe — to find that small niche of other humans that afford us some measure of human interaction while still permitting us to define ourselves as a Thing Set Apart. The screwball, the chili powder, the fancy windup, the obscure quotes about Catholicism from sociology Ph.D.s in your investing think-piece — instead of a barbaric yawp, it becomes a signal to your tribe. When pressed, our willingness to rip off the steering wheel and adopt a competitive strategy becomes dominant, a necessity. Lingering in the back of our heads as we go all-in on our tribe is the knowledge that our tribal leaders, no matter who they are, will sell us down the river every time.
In our investing lives, when we build portfolios, we know full well how many options our clients or constituents have, so these three competing impulses drive our behaviors: cynicism, affectation and tribalism. The cynical, nihilistic impulse shouts at us that nothing matters enough to justify risking being fired, and so we end up choosing the solution that looks most like what everyone else has done. That’s the ultimate equilibrium play we’re all headed toward anyway, right? The affectation impulse requires that we add a little something to distinguish us from our peers. A dash of chili powder. A screwball here or there, or an outlandish delivery to delight and astonish. Our tribal impulse compels us toward the right-sounding idea that makes us part of a group (I’m looking at you, Bogleheads). More frequently, we’re motivated by a combination of all three of these things in one convoluted, ennui-laden bit of arbitrary decision-making.
The real kick in the teeth of all this is that many of the things we are compelled to do by these impulses are actually good and important things, even Things that Matter. But because of the complex rationale by which we arrive at them (and other biases besides), we often implement the decisions at such a halfhearted scale that they become irrelevant. In other, worse cases, the decisions function like the tinkering we discussed in And They Did Live by Watchfires, potentially creating portfolio damage in service of a more compelling marketing message or to satisfy one of these impulses. In both cases, these flourishes and tilts are too often full of sound and fury, signifying nothing.
Too Little of a Good Thing
What, exactly, are we talking about? Well, how about value investing, for starters?
I think this one pops up most often as a form of the tribal impulse, although clearly many advisors and allocators use it as a way to add a dash of differentiation as well. Now, most of us are believers in at least a few investing tribes, each with its own taxonomy, rituals, acolytes and list of other tribes we’re supposed to hate in order to belong. But none can boast the membership rolls of the Value Tribe (except maybe the Momentum Tribe or the Passive Tribe). And for good reason! Unlike most investment strategies and approaches devised, buying things that are less expensive and buying things that have recently gone up in price can both be defended empirically and arrived at deductively based on observations of human behavior. The cases where science is really being applied to investing are very, very rare, and this is one of them. Rather than pour more ink into something I rather suppose everyone reading this believes to one extent or another, I’d instead direct you to read the splendid gospel from brothers Asness, Moskowitz and Pedersen on the subject. Or, you know, if you’re convinced non-linearities within a population’s conditioning to sustained depressing corporate results and lower levels of expected growth mean that such observations are only useful for analysis of the actions of an individual human and can’t possibly be generalized or synthesized into a hypothesis underpinning the existence of the value premium as an expression of market behavior, then don’t read it. Radical freedom!
What is shocking is how ubiquitous this belief is when I talk to investors, and how little investors demonstrate that belief in their portfolios. We adhere to the tribe’s religion, but now that it’s not a sin to skip out, we only attend its church on Christmas and Easter. And maybe after we did something bad for which we need to atone.
Value is the more socially acceptable tribe (let’s be honest, momentum has always had a bit of a culty, San Diego vibe), so let’s use that as our case study. Since I’m worried I’m leaving out those for whom cynicism is the chosen neurosis, let’s use robo-advisors to illustrate that case study. They’re instructive as a general case as well, since they, by definition, seek to be an industry-standard approach at a lower price point. Now, of the two most well-advertised robos, one — Wealthfront — mostly ignores value except in context of income generation. The other — Betterment — embraces it in a pretty significant way. I went to their very fine website and asked WOPR what a handsome young investment writer ought to invest in to retire around 2045. Here is what they recommended:
Source: Betterment 2017. For illustrative purposes only.
Pretty vanilla, but then, that’s kind of the idea of the robo-advisor. But I see a lot of registered investment advisors and this is also straight out of their playbook. It’s tough to find an anchor for the question “I know I want/need value, but how much?” As a result, one of the most common landing spots I see is exactly what our robot overlords have recommended: half of our large cap equities in core, and the other half in value. We signal/yawp a bit further: we can probably also afford to do it in the smaller chunks of the portfolios, too. Lets just do all of our small cap and mid cap equities in a value flavor. As for international and emerging equities, we don’t want to scare the client with any more line items or pie slices invested in foreign markets than we need, so let’s just do one big core allocation there.
I’m putting words in a lot of our mouths here, but if you’re an advisor or investor who works with clients and this line of thinking doesn’t feel familiar to you, I’d really like to hear about it. Because this is exactly the kind of rule of thumb I see driving portfolio decisions with so many allocators that I speak to. But how do we actually get to a portfolio like this? If you think there’s a realistic optimization or non-rule-of-thumb-driven investment process that’s going to get you here, let’s disabuse ourselves of that notion.
Could plugging historical volatility figures and capital markets expectations into a mean/variance optimizer get you to this split on value vs. core? In short? No. No, we know that this is an impossible optimizer solution because the diversification potential at the portfolio level — what we call the Free Lunch Effect in this piece — would continue to rise as we allocated more and more of our large cap allocation to a value style (and less and less to core). In other words, while the intuition might be that having both a core and value allocation is more diversifying (more pie slices!), that just isn’t true. In a purely quantitative sense, you’d be most diversified at the portfolio level with no core allocation at all!
Free Lunch Effect of Various Allocations to Large Cap Value vs. Large Cap Core in Example Portfolio
Source: Salient 2017. For illustrative purposes only.
If your instinct is to say that doesn’t look like much diversification, however, you’d be right as well. Swinging our large cap portfolio from no value to nothing but value reduces our portfolio risk by around 8bp without reducing return (i.e., the Free Lunch). That’s not nothing, but it’s damn near. The reason is that the difference between the Russell 1000 Value Index and the Russell 1000 Index or the S&P 500, or the difference between your average large cap value mutual fund and your average large cap blend mutual fund, is not a whole lot in context of how most things within a diversified portfolio interact. Said another way, the correlation is low, but the volatility is even lower, which means it has very little capacity to impact the portfolio. Take a look below at how much that value spread contributes to portfolio volatility. The below is presented in context of total portfolio volatility, so you should read this as “If I invested all 32% of the large cap portion of this portfolio in a value index and none in a core index, the value vs. core spread itself would account for about 0.1% of portfolio volatility.”
Percentage of Portfolio Volatility Contributed by LC Value-Core Spread
Source: Salient 2017. For illustrative purposes only.
Fellow tribesmen, does this reflect your conviction in value as a source of return? Some of you may quibble, “Well, this is just in some weird risk space. I think about my portfolios in terms of return.” Fine, I guess, but that just tells the same story. Consider how most value indices are constructed, which is to say a capitalization weighted splitting of “above average” vs. “below average” stocks on some measure (e.g., Russell) or multiple measures (e.g., MSCI) of value. We may have in our heads some of the excellent research on the value premium, but those are almost always expressed as regression alphas or as spread between high and low quintiles or deciles (Fama/French) or tertiles (Asness et al). In most cases they are also based on long/short or market neutral portfolios, or using methodologies that directly or indirectly size positions based on the strength of the value signal rather than the market capitalization of the stock. There are strategies based on these approaches that do capitalize on the long-term edge of behavioral factors like value. But that’s not really what you’re getting when you buy most of these indices or the many products based on them.(3)
So what are you getting? For long-only stock indices globally, probably around 80bp(4) and that assumes no erosion in the premium vs. long-term average. Most other research echoes this – the top 5 value-weighted deciles of Fama/French get you about 1.1% annualized over the average since 1972, and comparable amounts if you go back even further. Using the former figure, if you swung from 0% value to 32% value in your expression of your large cap allocation — frankly a pretty huge move for most investors and allocators — we’re talking about a 26bp difference in expected portfolio returns. Again, not nothing, but if our portfolio return expectations are, say, 8%, that’s a 3.2% contributor to our portfolio returns under fairly extreme assumptions.
Does this reflect your conviction in value as a source of return? No matter how we slice it, the ways we implement even fundamental, widely understood and generally well-supported sources of return like value seem to be a bit long on the sound and fury, but unable to really drive portfolio risk or return. Why is this so hard? Why do we end up with arbitrary solutions like splitting an asset class between core and value exposure like some sort of half-hearted genuflection in the general direction of value?
Because we have no anchor. We believe in value, but deep down we struggle to make it tangible. We don’t know how much of it we have, we don’t even know how much of it we want. We struggle even to define what “how much” means, and so we end up picking some amount that will allow us to sound sage and measured to the people who put their trust in us to sound sage and measured.
I’m going to spend a good bit of time talking about how I think about the powerful diversifying and return-amplifying role of behavioral sources of return like value as we transition our series to the Things that Matter, so I’ll beg both your patience and indulgence for leaving this as a bit of a resolutionless diatribe. I’ll also beg your pardon if it looks like I’ve been excessively critical of the fine folks who put together the portfolio that has been our case study. In truth, that portfolio goes much further along the path than most.
The point is that for various behavioral reasons, our style tilts like value, momentum or quality occupy a significant amount of our time, marketing and conversations with clients, and — by and large — signify practically nothing in terms of portfolio results. In case I wasn’t clear, yes, I am saying that value investing — at least the way most of us pursue it — doesn’t matter.
The Magically Disappearing Diversifier
The time we spend fussing around with miniscule style tilts, however, often pales in comparison to the labor we sink into our flourishes in alternatives, especially hedge funds. Some of this time is well-spent, and well-constructed hedge fund allocations can play an important role in a portfolio. When I’m asked to look at investors’ hedge fund portfolios, there are usually two warning signs to me that the portfolios are serving a signaling/tribal purpose and not some real portfolio objective:
- Low volatility hedge funds inside of high volatility portfolios that aren’t using leverage
- Hedge fund portfolios replacing Treasury or fixed income allocations
Because of the general sexiness (still, after all these years!) of hedge fund allocations to many clients or constituents, the first category tends to be the result of our affectation impulse. We want to add that low-vol, market-neutral hedge fund, or the fixed income RV fund that might have been taking some real risk back in 2006 when they could lever it up a bajillion times, not because of some worthwhile portfolio construction insight, but perhaps because it allows us to sell the notion that we are smart enough to understand the strategies and important enough to have access to them. Not everyone can get you that Chili P, after all. In some cases, sure — we are signaling to others that we are also part of that smart and sophisticated enough crowd that invests in things like this. In the institutional world, where it’s more perfunctory to do this, it’s probably closer to cynicism: “Look, I know I’m going to have a portfolio of low-vol hedge funds, so let’s just get this over with.”
For many clients and plans — specifically those where assets and liabilities are mostly in line and the portfolio can be positioned conservatively, say <10% long-term volatility — that’s completely fine. But for more aggressive allocations, there is going to be so much equity risk, so much volatility throughout the portfolio, that the notion that these portfolios will serve any diversification role whatsoever is absurd. They’re just taking down risk, and almost certainly portfolio expected returns along with it. Unless you feel supremely confident that you’ve got a manager, maybe a high frequency or quality stat arb fund, that can run at a 2 or 3 Sharpe, it is almost impossible to justify a place for a <4% volatility hedge fund in a >10% target risk portfolio. They just won’t move the needle, and there are better ways to improve portfolio diversification, returns or risk-adjusted returns.
The second category starts to veer out of “Things that Don’t Matter” territory into “Things that Do Matter, but in a Bad Way.” More and more over the last two years, as I’ve talked to investors their primary concern isn’t equity valuations, global demographics, policy-controlled markets, deflationary pressures, competitive currency crises, protectionism, or even fees! It’s their bond portfolio. The bleeding hedge fund industry has been looking for a hook since their lousy 2008 and their lousier 2009, and by God, they found it: sell hedge funds against bond portfolios! Absolute return is basically just like an income stream! There seems to be such a strong consensus for this that it may have become that cynical equilibrium.
No. Just no.
It’s impossible to overstate the importance of a bond/deflation allocation for almost any portfolio. This is an environment that prevails with meaningful frequency that has allowed the strong performance of one asset historically: bonds, especially government bonds (I see you with your hands raised in the back, CTAs, but I’m not taking questions until the end). The absolute last thing any allocator should be thinking about if they have any interest in maintaining a diversified portfolio, is reducing their strategic allocation to bonds. I’ll be the first to admit that when inflationary regimes do arrive, they can be long and persistent, during which the ability of duration to diversify has historically been squashed. The negative correlation we assume for bonds today is by no means static or certain, which is one of the reason I favor using more adaptive asset allocation schemes like risk parity that will dynamically reflect those changes in relationship. But even in that context, the dominance and ubiquity of equity-like sources of risk means that almost every investor I see is still probably vastly underweight duration.
Now many of us do have leverage limitations that start to create constraints, and so I won’t dismiss that there are scenarios where that constraint forces a rational investor not to maximize risk-adjusted returns, but absolute returns. I’m also willing to consider that on a more tactical basis, you may be smarter than I am, and have a better sense of the near-term direction of bond markets. In those cases, reducing bond exposure, potentially in favor of absolute return allocations, may be the right call. But if you have the ability to invest in higher volatility risk parity and managed futures, or if you have a mandate to run with some measure of true or derivatives-induced leverage, my strong suspicion is that you’ll find no cause to sell your bond portfolios in favor of absolute return.
Ultimately, it’s hard to be too prescriptive about all this, because our constraints and objective functions really may be quite different. To me, that means that the solution here isn’t to advise you to do this or not to do that, except to recommend this:
Make an honest assessment of your portfolio, of the tilts you’ve put on, and each of your allocations. Do they all matter? Are you including them because of a good faith and supportable belief that they will move the portfolio closer to its objective?
If we don’t feel confident that the answer is yes, it’s time to question whether we’re being influenced by the sorts of behavioral impulses that drive us elsewhere in our lives: cynicism, affectation and tribalism. In the end, the answer may be that we will continue to do those things because they feel right to us and our clients. And that may be just fine. A little bit of marketing isn’t a sin, and if your processes that have served you well over a career of investing are expressed in context of a particular posture, there’s a lot to be said for not fixing what ain’t broken. There’s nothing wrong with an impressive-looking windup, after all, until it adversely impacts the velocity and control of our pitches.
What is a sin, however, is when a half-hearted value tilt causes us to be comfortable not taking advantage of the full potential of the value premium in our portfolios. When the desire to get cute with low-vol hedge funds causes us to undershoot our portfolio risk and return targets. Perhaps most of all, when we spend our most precious resource — time — designing these affectations. We will be most successful when we reserve our resources and focus for the Things that Matter.
(1) Please – no letters about his relief starts in 1980. If MLB called him a rookie, imma call him a rookie.
(2) Probably the only exception in this conversation is Randy Johnson, who, while mostly vanilla in his mechanics, would probably get feedback from a coach today about his arm angle, his hip rotation and a whole bunch of other things that didn’t keep him from striking out almost 5,000 batters.
(3) As much as marketing professionals at some of the firms with products in this area would like to disagree and call their own product substantially different, they all just operate on a continuum expressed by the shifting of weightings toward cheaper stocks. Moving from left to right as we exaggerate the weighting scheme toward value, the continuum basically looks like this: Value Indices -> Fundamental Indexing -> Long-Only Quant Equity -> Factor Portfolios
(4) Simplistically, we’re just averaging the P2 and half of the P3 returns from the Individual Stock Portfolios Panel of Value and Momentum Everywhere, less the average of the full universe. An imperfect approach, but in broad strokes it replicates the general half growth/half value methodology for the construction of most indices in the space.
PDF Download (Paid Subscription Required): http://www.epsilontheory.com/download/16031/