The "received wisdom" about how to invest - on the equity side, at least - hasn't worked very well as a method for outperforming the benchmark Standard & Poor's 500 Stock Index, at least not for us at Federal Express Corp.
And, scanning the universe of other plans, it seems apparent the standard paradigm for large plan equity management hasn't worked for others. The trend to passive investing is proof enough of that.
Most large plans, influenced by academic notions of manager style, do not attempt to outperform the index by selecting managers that can evaluate the traditional sector and industry groupings, and the individual equities within those groupings. Rather, managers are selected and assets are allocated on the basis of "style."
We can find no real evidence that the "style" concept in equity portfolio management as applied to large capital pools actually works. It seems to us, the "specialist" approach can provide reliable excess returns over the index.
Managers who focus on growth stocks - or value stocks or yield stocks - are sometimes holding the stocks that lead the market upward, and sometimes not. Performance depends primarily on which style is in favor - perhaps more than manager skill.
"Style" management has become an entrenched approach to large-cap investing even though it is quite clear to all that a given style is likely to underperform the index for long periods of time.
It has become so entrenched that market forces being what they are, it has become almost impossible to find a manager who does not purport to adhere to one style or the other. This means you're stuck with dividing up your assets along style lines, knowing in advance some of those assets are going to underperform, hoping the others will outperform to a greater extent and further hoping that at the end of the day when the dust has settled, you'll come out ahead.
Before the advent of the concept of "styles," the stocks that constitute the index were subclassified according to their economic sector and industry group. Clearly, in both the real economy and in the stock market, the sector in which a company resides is a more telling prognosticator of success or failure than whether it is categorized as a growth stock or value stock. And, as most of us know, today's growth stock may become tomorrow's beaten-up value issue, and vice versa.
The irony in the "style bound" conceptualization of a plan is that it ignores the classifications that have stood the test of time: economic sectors and industry groups.
It is no accident that performance of sectors and industry groups has been reported since the inception of markets, while statistics for "style" performance are published infrequently, with limited distribution, and with arguable if not dubious component elements. Stocks are most logically categorized based on the nature of their business and the broader economic factors that more or less intensely affect it. "Styles" are vagaries, and that has been reflected in manager performance.
Nearly everyone agrees market timing doesn't work, at least in the large plan context. Styles have proven a roll of the dice. Yet it's clear managers often have a high level of expertise in distinguishing the true stand-out stocks from the poor ones, we see it all the time in the portfolios. But we also see the absence of representation in a hot sector can kill a good manager's relative performance, as can overweighting a sector that does poorly.
Why don't we eliminate sector betting by limiting each manager to only one sector, just as we have basically eliminated market timing by mandating full investment? Why don't we just hire managers to pick stocks, from a limited universe, restricting the playing field so there will be no "wrong sector" excuse, and no utterance, heaven help us, ever again, not even once, of those dreaded words, "our style was out of favor." What if we would just get a manager's best stock picks, the wheat without the chaff?
For us at Federal Express' defined benefit plan, either this concept had merit or we'd throw in the towel, and index. Would it work? Historical data was limited, but we found as much as we could. We examined our managers' performance, isolating the sectors in which they consistently outperformed, and those in which they didn't. (As you might expect, our managers who made money with the likes of Coke and Liz Claiborne came in limping when they tried their hand with International Paper and Alcoa). We put together a back-tested model portfolio using sector mutual funds - a rough approximation, because there really aren't any that purely follow the S&P sector breakdowns. We crunched the numbers and we were intrigued: it appeared possible to obtain excess performance over the S&P 500 while still retaining the same valuation and volatility characteristics as that benchmark. And we weren't seeing just a 50 basis point advantage, either. In one 12-month period, our patched-together backtest showed a possible 1,000 basis points of added value.
We had enough evidence, and we felt the idea was logical enough, to undertake a test. Instead of distributing our next contribution among the active managers as usual, we funded the inception of our Sector Plus program.
Starting with about $50 million, we used a combination of our own screening of our managers' hidden specialties and a questionnaire asking them which sector they'd like to undertake. Most got their first or second choice. Their mission was to outperform their sector. Not the whole market. Not their "peer group" (of which there was none). Not the INDATA universe. Just the 40 to 60 stocks in their sector. Surely, there will be one or two or three issues they can overweight to beat the sector, we thought. If each manager adds only 20 basis points to his sector, we concluded, our cumulative added value will be terrific.
Based on group meetings we held with the managers, we set up guidelines and limitations:
Each manager is responsible for one of the S&P sectors;
no more than 5% cash is allowed;
Options may be written;
Futures are not a part of the program;
Managers may use stocks that are not in the S&P 500 but are included in their sector (by SIC code) up to 30% of total portfolio weight;
American depository receipts may be used; and
No particular stocks are required to be held.
Aside from those few guidelines, we wanted to leave matters up to the manager. If a manager wanted to hold a three-stock portfolio, so be it. Our office and our consultants would monitor compliance monthly, and we set up a new portfolio analysis system so we soon were able to monitor compliance on a daily basis.
What we found was fascinating and, on its face at least, addressed many of the problems we'd been having as the stablemaster of a group of active managers. Bear in mind the sector managers were the same as our active "regular" managers; their sector accounts were separate and different accounts.
Whereas the active managers held about 260 stocks in the aggregate in their regular accounts, the sector managers held about the same number of stocks, but each held different ones. That was a far cry from the regular accounts, where more than half of the stocks were held by more than one manager, and where all eight held Philip Morris (is it value or is it growth?). At least we no longer were paying many managers for owning the same stock.
Too, because selected individual issues tend to dominate their sectors in terms of weight, we were automatically heavily represented in the key index stocks that truly move the S&P 500 - so even if performance was off, it could never be too far off.
Indeed, the statistical profile of our Sector Plus portfolio proved to be very much like that of the S&P 500. In all of the usual measures - capitalization, r-square, yield, alpha, beta - you'd think we were running an index fund with a 260-stock surrogate. Most students of portfolio management would insist it's not possible to beat an index with a portfolio whose attributes are virtually the same as the index, and most of our managers offered the same opinion. After all, you have to take more risk to get more reward, right?
Apparently not. For the first full year of our program, Sector Plus outperformed the index by 370 basis points. This was from Feb. 28, 1991, (just at the end of the Desert Storm rally) through Feb. 28, 1992, a time when the market rotated between value and growth quite notably and rapidly: most managers had a chance to say "our stocks are out of favor" on a number of occasions. During this period our managers did well in their individual "normal" accounts, but the aggregate of the active managers - the same ones managing in Sector Plus - still underperformed the index while Sector Plus was landing in the top quartile with no more risk than the overall market.
Since its inception, Sector Plus has outperformed the median equity manager, and has added 2.2% a year net of fees.
More importantly for us, Sector Plus also once again has outperformed the aggregate of our active managers.
In other words we're getting the kind of excess performance from active managers that we've wanted - which justifies having active managers in the first place - by providing a structure in which their stock-picking abilities can become the primary variable. Although portfolio attributes are much like the index overall, to the extent that managers are right in their stock picking our performance will be much greater than the index, because our likelihood of including market winners and excluding market losers is greatly enhanced. Our risk and volatility is actually lower than the aggregate of our active managers' "normal" accounts, although returns are much higher.
The theory of using many managers each with their own style relies on a kind of mystic faith that the right mix of spices will result in a tasty investment meal, but we and most others have experienced, instead, the kind of transient satisfactions that leave you hungry an hour later. As Mill put it: "Much time and effort is expended by many, but too often it is spent going off in numerous individual directions instead of developing a team of specialists dedicated to outperforming the index, each specialist with a realistic and focused part of the whole - rather than just hoping it will work out that way with a bunch of style managers."
At Federal Express, we've targeted 75% of our equity program to Sector Plus. Managers are given a manageable universe with which to work, performance is always relative to that abundantly clear sector benchmark, and the person who said "our style is out of favor" was last seen riding the "down" elevator.