评估业绩记录——案例研究

Memo to:

Oaktree Clients

From:

Howard Marks

Re:

Assessing Performance Records – A Case Study

What are the non-negotiable requirements for accurately assessing investment performance? I’d say:   

a record spanning a significant number of years, a period that includes both good years and bad, enabling us to assess performance under a variety of circumstances, and a benchmark or peer universe that makes for a relevant comparison.

The other day, at an event for alumni and other constituents of the University of Pennsylvania, president Amy Gutmann reviewed the performance of the university during the financial crisis. In the process, she had some kind words for Penn’s Investment Board, which I chaired for the ten years from June 30, 2000 through June 30, 2010. Thinking about it afterward, I realized that I should share with you the story of Penn’s endowment and its lessons. Penn has agreed that I may do so. The data is a little out of date, but the lessons aren’t.

A Little Background For roughly two decades starting in the late 1970s, Penn’s endowment was led by John Neff, probably the most respected investor of that era, in strict adherence to the principles of value investing. Thus, its fortunes fluctuated along with the performance of that school of thought. It did extremely well in the 1980s, when value beat growth and Penn outperformed the value school, and it lagged a little in the early ’90s, when those trends reversed. Penn’s portfolio completely lacked exposure to growth stocks, tech stocks, buyouts or venture capital. Thus it fell behind substantially in fiscal 1995 through fiscal 1999, when it gained 16% a year but the average of its peers gained about 23%. Then, in fiscal 2000, the last year before I became chairman, value stocks stagnated, tech stocks soared, and returns on venture capital ranged well into triple digits. Penn lost 2% while its peers averaged returns in the twenties and those with investments in the right venture capital funds made twice that. Penn constituencies such as alumni/donors and the administration were very unhappy in those years, despite the endowment’s high absolute overall return. Penn appointed its first Chief Investment Officer in the late 1990s and began to diversify the portfolio beyond value. Having been on the Investment Board for a few years, I became its chairman at the start of fiscal 2001. The events of the next decade, along with the decisions made and actions taken, provide a number of valuable lessons.

2012 Oaktree Capital Management, L.P.

All Rights Reserved

What Matters? The first of those lessons is that the ability to ignore relative performance depends on the circumstances and, in particular, the constituencies the performance has to please. Endowments provide an outstanding example of this phenomenon. Intellectually, no one could be unhappy with 16% a year for five years. In fact, the trustees of a private foundation probably would have been delighted with such a return in FY1995-99. But it’s harder for an institution with outside constituents, like a university, to dismiss relative performance. Alumni question the management of the endowment, and prospective donors start to say, “I love the school, but it makes no sense for me to make my gift now. I’ll hold onto the money, grow it at a rate above what you’re achieving, and give it later.” No university president wants to be the recipient of that message. In order to survive and have a chance to produce long-term performance, investors have to live up to their constituents’ expectations in the short run. Of course, it’s important to inculcate reasonable expectations, or to choose clients who have them. But ultimately, the manager’s job isn’t to make money, it’s to deliver client satisfaction, so expectations have to matter. All of us, on both sides of the process, should be sure we know what pattern of performance is expected. How else can we know how satisfaction can be delivered? As a result of holding a highly idiosyncratic portfolio, Penn experienced performance that deviated – unfavorably – from that of its peers to an extent that became intolerable. This necessitated change.

Do It Now? Upon starting in on the job, I was immediately confronted by one of the truly classic investment dilemmas: You take on the management of a portfolio, and you just know it’s structured wrong in principle. In Penn’s case, it was clearly unwise – probably in terms of optimizing risk and return, and certainly in terms of keeping up with peers, and thus expectations – to completely omit the things that had been excluded from Penn’s portfolio. I knew right away that Penn’s portfolio should include some exposure to growth, tech, buyouts and venture capital. But the reason their exclusion had become so painful is that they had done so well for a half-decade. So in principle you should own something, but its price is skyhigh. Should you hold your nose and buy at what may be excessive prices? Or should you wait for a correction, at the risk of continuing to underperform if it goes higher (since we know how often things that are overpriced can continue upward)? Whenever I’m presented with this dilemma, I trot out a 1957 cartoon from The New Yorker Magazine that was reproduced in the Financial Analysts Journal in 1975. It’s my absolute favorite, and I’ve been waiting for an opportunity to share it with you:

2012 Oaktree Capital Management, L.P.

2

All Rights Reserved

For me, this cartoon frames the question precisely: Should we do the thing that’s right in principle, or should we alter our behavior to reflect today’s real-world conditions? There’s no one right answer; it goes back to expectations. If it’s important to track the competition, you should start to make the portfolio less idiosyncratic, regardless of price attractiveness. But if you care more about absolute performance, achieved with risk under control, you should refuse to buy sky-high assets. The latter is my preference, and it was reflected in my decision. Penn held off from buying tech and growth stocks. But I had declared my intention before taking the job. To put it in horribly mixed metaphors, having missed the boat for six years, I said I wouldn’t jump on the bandwagon just in time to ride it over the cliff. I hoped making this clear would condition expectations and provide cover in case our actions initially proved wrong.

Offense or Defense? The dilemma just discussed relates to the biggest single issue facing anyone tasked with structuring a portfolio: whether to stress offense – trying for high returns – or defense – reducing the likelihood of losses. Of course most investors balance these two things. The question is in what proportion.

2012 Oaktree Capital Management, L.P.

3

All Rights Reserved

Penn’s historic emphasis on value investing and its eschewing of bigger potential money makers had a lot to do with defense, especially in the environment of the late 1990s. Likewise, I believe I am known – and I certainly know myself – to be one who usually puts great emphasis on defense. Would a cautious approach continue to penalize Penn, or was it what was called for under the circumstances? Having fallen so far behind, should we continue to stress defense to avoid losses if the market reversed course, or should we go on the offensive in an attempt to make up the lost ground? This question had particular importance at Penn. Given its early history as a commuter school rather than an elite institution like some of its peers, Penn came into the 21st century underendowed; it ranked only 70th in the country in endowment per student. So the stewards of Penn’s endowment faced a particular dilemma: should we invest conservatively because we can’t afford to lose the little bit we had, or aggressively in an attempt to close the gap? Again, there’s no one right answer to that question, and perhaps there was no one right answer for Penn. But the answer was clear for me: I wouldn’t preside over a shift to offense . . . and especially not on the heels of one of the best decades for stocks in history. Working with CIOs Landis Zimmerman (now at Howard Hughes Medical Institute) in the early years and especially closely with Kristin Gilbertson in 2004-2010, the Investment Board and I led gradual diversification into growth stocks, emerging markets and defense-oriented hedge funds, with an emphasis on managers stressing risk-control. We established an allocation for private equity but implemented it very slowly. We kept an above-average percentage of the portfolio in publicly traded securities. And, importantly, we maintained a substantial allocation to cash and U.S. Treasurys, solely to enable us to meet the need for cash for operations and thereby avoid having to sell assets in a time of depressed prices.

The Results The performance produced by these decisions was quite predictable. With its low-risk portfolio, Penn outperformed when risk taking was penalized but trailed when risk taking was rewarded. It outperformed when value stocks did well but lagged when more aggressive tools, including leverage and portable alpha, paid off. For the decade overall it lagged the average of its peer institutions by a small margin and exhibited lower volatility. No surprise there. Penn’s return was about 5½% for FY2001-10, while most of its peers made 6% or 7%. But average results don’t tell the whole story. It’s important to remember one of my favorite adages, about the six-foot-tall man who drowned crossing the stream that was five feet deep on average. In investing, it’s not enough to survive “on average.” You have to survive on the worst days, when the low points in the market are reached. My decade as chairman provided an outstanding opportunity to see that adage in action.

2012 Oaktree Capital Management, L.P.

4

All Rights Reserved

The Realities of Risk and Return In late 2008 and early 2009 (in other words, for universities, fiscal year 2009), the global financial crisis presented the greatest sinkhole in eighty years. Those caught mid-stream without life jackets were penalized. Many of Penn’s leading peer institutions lost 25-28% that year, while Penn’s loss was “only” 15½%. I described Penn’s results, loosely speaking, as “the least worst.” Going from the investment arena to the real world of university operations makes it clear that investment risk isn’t an abstraction. No, risk isn’t just volatility. It’s what happens to owners of capital when downward fluctuations occur and principal losses are experienced. Many of Penn’s peers were forced to curtail some of their spending, ranging from hot breakfasts to student aid. Some had to suspend construction projects. There were freezes on hiring and wages. Some put illiquid partnership interests up for sale to raise cash and/or escape continuing funding obligations. And some had to borrow in the taxable bond market to meet cash needs. Penn, on the other hand, had lots of liquidity and faced little in the way of capital calls. Thus it didn’t have to go on the defensive operationally. Instead, it was able to keep hiring faculty, keep giving grants instead of loans, and take advantage of an attractive opportunity to purchase adjacent acreage. The benefits of risk control were made concrete. And the performance of the endowment, the CIO and the Investment Board – and, as I said earlier, yours truly – became the subject of some very nice words.

What If? – Part I All of the above is history. It presented taxing dilemmas and important choices, but other than as to degree, nothing portfolio managers, CIOs and investment committees don’t face routinely. But there are two hidden issues – both somewhat philosophical – that I find far more interesting, provocative and important. They surround questions of timing and chance. Sometimes investors feel something is going to happen in the period ahead, and that they should do something about it in their portfolios. And sometimes they’re right. But rarely do the anticipated events occur as expected, and thus rarely are investors’ actions proved correct immediately. Overpriced assets continue to appreciate, and cheap stocks decline further. Even if they do the right thing, very few investors do it at just the right time. Thus timing – and in particular the selection of the beginning point and end point for studying a performance record – plays an incredibly important role in perceptions of success or failure. In his important book Fooled by Randomness, Nassim Nicholas Taleb points out how easily random events can make good decisions look wrong and bad decisions look right. Clearly one of the reasons for this is that events don’t happen on schedule. I think of investment performance as what happens when events collide with a pre-existing portfolio. A good decision can wisely anticipate the range of things that might happen, but the one thing that actually happens may not have been one of the ones that reasonably could have been considered highly likely

2012 Oaktree Capital Management, L.P.

5

All Rights Reserved

– or even possible. Or it may just happen at a time other than when it “should” have. The bottom line is that investors are often “right for the wrong reason,” and vice versa. So the first key observation is this: I came into the job at the end of a highly bullish period, maintained a cautious approach, and it worked. What if I’d gotten the job two years earlier? I probably would have done many of the same things. But now my tenure would have included the horrendous FY2000, with Penn’s 3,000 basis point underperformance. And it would have omitted FY2009, in which Penn lost 1,200 basis points less than many peers and avoided being hamstrung. If I had led Penn’s endowment to take the same actions in FY19992008 as it did in FY2001-2010, which is quite likely, I’d be considered a very average chairman . . . at best. The principle lesson of this tale is the observation that investment timing is imprecise and difficult but extremely significant in terms of outcomes. The bottom line: be understanding when evaluating track records, and refuse to accept the results at first glance. Taleb reminds us to wonder about “alternative histories” – the other things that reasonably, probably could have happened. Doing so isn’t easy, but it’s essential in any field in which randomness plays a big part.

What If? – Part II The second philosophical question is similar but even simpler: what if the global financial crisis hadn’t occurred? My tenure as chairman of Penn’s endowment was marked by my characteristic preference for being able to survive bad times over acting to take maximum advantage of good times. And the worst financial event in eighty years materialized, making that the right approach for the times. But was I “right”? As I said before, during my tenure, Penn underperformed by a bit: 5½% versus 6-7%, a very reasonable sacrifice in exchange for side-stepping the pain of the crisis. But if we take out FY2009, the results for the other nine years were 8% for Penn versus 10½-11½% for its peers. Is that still a reasonable sacrifice? It was largely the arrival of the crisis – with its influence on the ten-year record and its highly visible impact on university operations – that made my tenure a successful one. But could the crisis reasonably have been anticipated, making my caution appropriate? Or was it unforeseeable and thus fortuitous, meaning I was right for the wrong reason? I wonder about this a lot, in order to derive the significance of my tenure as chairman and learn from it. (If you want to read more about “what if” questions like these, you might enjoy the section called “What’s Real?” in my memo Pigweed, from December 7, 2006.)

2012 Oaktree Capital Management, L.P.

6

All Rights Reserved

The conclusions of my introspection are as follows:      

I lean strongly toward investing defensively unless asset prices are so low and investors so chastened that less cautious behavior is called for. In Penn’s position endowment-wise, an emphasis on defense was appropriate. The euphoric market behavior of the late 1990s, and especially 1999, made elevated caution particularly compelling. It was certainly the events that unfolded that made my actions seem as right as they do. The events marking the crisis came largely as a surprise. Although I was increasingly worried in the period from late 2004 through the first half of 2007, I did not foresee the specific events of the global financial crisis of 2008, or its severity. Although Penn’s approach was generally cautious throughout, we increased the defensiveness of the portfolio significantly in 2006-08. Thus it can’t be argued that the portfolio stayed equally cautious all the time and eventually was bailed out by the crisis.

The bottom line – however you slice it – is that a cautious approach was appropriate under the circumstances, and it paid off for Penn. No strategy works all the time, but defensiveness was right for Penn as things turned out. Lucky or good? It’s always hard to tell.

*

*

*

A full ten years, but still the results were obviously highly dependent on the vagaries of timing and on some largely unforeseeable events. Were the actions taken at Penn right? Wrong? Or right for the wrong reason? We should insist on engaging in this kind of examination. Only then can we draw reliable conclusions and hope to improve our decision making. Let me know if Oaktree can help in this regard.

February 15, 2012

2012 Oaktree Capital Management, L.P.

7

All Rights Reserved