What this covers
Patrick interviews David Salem, an allocator and manager-selector, about how institutional portfolios get built and how to judge the people running them. The history comes first: legal opinions limiting institutions to investment-grade bonds forced sales of fallen credits, an opening David Swensen used at Yale, while Jack Meyer used Harvard's credit standing to borrow cheaply and fund an internal alpha engine of fixed income arbitrage. Harvard's trouble in 2008 and early 2009 becomes a case for specifying liquidity under worst-case conditions rather than only a drawdown tolerance.
The portfolio-construction section is where Salem turns contentious. He calls infrastructure and hedge funds contractual arrangements rather than asset classes, argues that under stress most holdings reduce to equity beta or sovereign credit, and predicts convergence on two or three buckets. He finds the Endowment Model too costly, complex, and opaque, and endorses Seth Klarman's view that cash should be the default. He also cites Peter Bernstein on diversification and describes distributional thinking. His Asia ex-Japan approach rests on prices and family-controlled firms, since fast-growing economies have tended to produce subpar stock returns.
The last stretch concerns evaluating managers. Salem favors timestamped purchase memos for ex-ante versus ex-post evaluation. He treats single-scenario commentary and poor writing as warning signs, doubts that fees can fully align interests, and says Regulation FD has pushed investors toward judging personality and culture. His central test is whether a manager has consciously pursued excellence.
Salem argues that successful asset allocation depends on the investor's mindset (ego-free, willing to be wrong and alone, pursuing excellence) more than on copying the strategies of the Yale or Endowment Model, and that manager evaluation should weigh character and culture over process.
- Institutions that copied Yale's tactics without its mindset missed the point, and the Endowment Model became overused through return-chasing
- Ego is insecurity, and insecurity shows up in gratuitous self-promotion, evasiveness about capacity, and departure from stated process
- Diversification is the only rational deployment of ignorance, and the best allocators plan for a distribution of outcomes and worst-case liquidity
Ego and insecurity revealed in conduct are the key test of managers.
- The most essential attribute to look for in a manager is evidence across their life that they consciously pursued excellence in serious endeavors, even if they never achieved it, because they judge themselves daily by whether they are doing their best.
“do I see evidence that they've consciously pursued excellence in just about everything they've done?”
- In Salem's working definition, ego equals insecurity, so the useful measure is where a manager sits on a spectrum from insecure self-display to graceful, secure conduct under pressure.
“in my own lexicon, which is maybe not everyone's, ego equals insecurity”
- Interviewers can probe a manager's history with open questions about sports, high school, college major choice, mentors, and first job to find evidence of conscious pursuit of excellence, though the method is highly imperfect.
“What sports did you engage in? What were you like as a high schooler? How did you choose your major in college?”
- The single best tell of insecurity is a gratuitous, proactive offering of thoughts, observations, and connections meant to impress, because that behavior is synonymous with insecurity.
“the single best tell in my judgment ... is to engage in a gratuitous and proactive offering up of thoughts, observations, and connections”
- Armstrong's behavior at a dinner, including arriving late and name-dropping Tiger Woods and Robin Williams while ignoring his guests, revealed extreme insecurity that Salem took as evidence of deception, consistent with attacking critics.
“the insecurity that was being telegraphed through that behavior was so stunning and profound that I thought there's no way anyone could've achieved rightly and legitimately what he professes to have achieved”
- The Mount Everest question, asking what one feat you would accomplish if no one could ever know, reveals character and is a useful lens on ego because most people reveal their motives in five to fifty seconds.
“if you could accomplish just one feat, F-E-A-T, in your lifetime subject to the condition that no one would ever know about it. What would it be?”
Screen managers on character, culture, incentives, and clarity before considering process.
- Salem evaluates money managers, and people generally, with a four-part framework that starts by screening out disqualifying attributes, then unfavorable ones, then looks for favorable attributes.
“It's a four part framework and it starts with disqualifying attributes and then moves to unfavorable and then moves to favorable.”
- A manager whose commentary assumes one unfolding scenario, rather than a distribution of probable outcomes, is a disqualifying trait because the world does not work in single scenarios.
“Single scenario thinkers are very dangerous in our businesses.”
- Maximizing a manager's personal financial stake is a misguided focus, because a manager might pledge their GP interest to charity and be motivated by psychological rather than pecuniary incentives, which can be preferable.
“I actually think that's a misguided focus.”
- Investors cannot know in advance when a manager's current process will become obsolete, so evaluating personality and culture, meaning whether the team treats money management as a profession willing to return capital, matters more than the process itself.
“because you cannot ex-ante the date on which the process they employ today when you allocate money to them will become obsolete”
- An allocator cannot apply one fee template to all managers and asset classes; in some cases a simple flat fee works, as with Andrew McDermott, who has most of his net worth invested alongside clients and so has strong pecuniary and psychological incentives.
“Just keep it simple. Why would you do that? Because Andrew has almost his entire net worth”
- Regulation FD and its enactment have profoundly changed communications between public company management and outside constituencies, making it very difficult for outside investors to assess where a company is going, which pushes investors toward judging personality and culture.
“particularly after Reg FD and the profound changes that its enactment by Congress has affected”
- A manager who is unwilling to specify how large an asset base their strategy can profitably absorb is often reason enough to end the conversation quickly.
“unwilling to specify those limits is often reason enough to terminate a conversation quite quickly”
- A carefully written, timestamped memo recording the reasons for a purchase before the outcome is known allows ex-post evaluation of whether a manager was right for the right reasons, which helps distinguish skill from luck.
“what if there's an ex-ante explanation that's written out very carefully ahead of time”
- Clear writing and clear thinking are synonymous, so a manager who writes poorly is unlikely to be a clear thinker, even though many people disagree.
“I think that clear writing and clear thinking are synonymous.”
- Peter Bernstein critiqued Salem's quarterly letters with typed, one-to-two-page responses over a decade, which Salem regards as the greatest professional kindness because it helped him think more clearly.
“a critique of my immediately prior quarterly letter. Often a page, sometime a page and a half.”
Diversify across few buckets because the future cannot be known.
- Overly diversified portfolios like the Endowment Model's are too costly, complex, and opaque, so the answer to uncertainty is diversification with fewer buckets, not maximal diversification.
“These portfolios are way too costly and complex and opaque.”
- Peter Bernstein's line that diversification is the only rational deployment of our ignorance is Salem's favorite investing principle, because we cannot know how the future unfolds.
“Diversification is the only rational deployment of our ignorance.”
- Infrastructure is not an asset class and hedge funds are not an asset class but a series of contractual arrangements, and many such labels are marketing schemes masquerading as asset classes.
“There's infrastructure. It's not an asset class. Obviously hedge funds is not an asset class. It's a contractual arrangement.”
- Institutional portfolios are converging on three generic buckets (total return, hedging, diversifying strategies) and sometimes two, because many purported asset classes are marketing schemes and in worst-case conditions most assets reduce to equity beta or sovereign credit.
“there really aren't 15 different asset classes in the world. There's kind of stuff that's tied to equity beta, there's stuff that tied to sovereign, full faith in credit”
- Cash should be the default asset rather than the S&P 500 or a 60/40 portfolio, as Seth Klarman argued, so allocators should shift into riskier assets only when a specific opportunity justifies it.
“Cash should be the default asset, not the S&P 500.”
Allocator judgment depends on a willingness to be wrong and alone.
- Managers and institutions trying to replicate Yale's success mimic its strategies and tactics rather than its underlying mindset of being willing to be wrong and alone, which is why the Endowment Model became overused by chasing recent returns.
“people have tried to mimic the strategies and the tactics as opposed to the underlying mindset and philosophy”
- Swensen's success at Yale came from his personality and mindset more than from the shift toward equities and absolute return funds, consistent with Emerson's idea that great institutions are the shadow of one person.
“It was the person. It was the man, if you will.”
- Successful allocators need a mindset willing to be wrong and alone, which means asking sincere questions even when answers embarrass them, since no question about how markets operate or managers deploy capital should be off limits.
“You have to have a mindset that says, 'I am willing to be wrong and alone.'”
- Marathon's co-founders persuaded the TIFF board because their presentation matched the coherent investment thinking they had written about in advance, including capital cycle analysis, and displayed ethics and a professional attitude.
“What they said tied in with what they'd written about in advance. It was just a coherent whole.”
- Accumulated experience and pattern recognition, like Dr. John Jane's immediate diagnosis of a shattered elbow, cannot be easily displaced by AI and machine learning, since such wisdom still matters even in medicine.
“the accumulated wisdom and reservoir of experience that caused John Jane that morning ... to come in and apply just common sense.”
- To earn the discretion to deploy capital in ways that may look wrong and alone, a CIO must build a reservoir of confidence and goodwill with the investment committee and board through genuine persuasiveness, not boastful self-promotion.
“you need to be persuasive. And I don't mean persuasive in kind of a Trump-like way”
Institutional models succeed only when leverage, capacity, liquidity, and taxes fit.
- Jack Meyer at Harvard realized the university had an extraordinarily high credit standing, so he could borrow at very low cost from banks that did not track Harvard's aggregate exposure, funding an internal alpha engine built on leverage.
“Harvard university, at least then, had the highest credit rating in the world.”
- A manager who has not studied how much similar capital is chasing the same strategy, and how much might flow in over the relevant horizon, has not thought rigorously about how capacity affects returns, which Salem calls the fallacy of composition.
“An insensitivity to external changes in the external environment. My writing's in a different context with respect to asset allocation, call that the fallacy of composition.”
- Harvard's internal alpha engine under Meyer relied on fixed income arbitrage such as on-the-run versus off-the-run Treasury convergence trades, which were very profitable with substantial leverage and sound risk management.
“Simple things like on the run versus off the run treasuries, where you thought there would be over time convergence”
- A drawdown tolerance alone is insufficient; investors must specify liquidity demands under worst-case conditions before taking on risk, and Harvard's failure to do so contributed to its problems in 2008 and early 2009.
“You need to be crystal clear up front about what are the liquidity demands on the portfolio under worst case conditions.”
- Chuck Feeney's giving was anonymous, and DFS's excess profitability was kept secret because public knowledge would have led competitors to arbitrage away the concession-based returns.
“if people knew, then the excess return would've been competed and arbitraged away.”
- Applying the endowment model to substantial taxable private fortunes is a big mistake because taxes sit between the cup and the lip, creating drag that endowments, which are tax-exempt, do not face.
“And it's a big, big mistake principally because of the interposition of taxes between the cup and the lip”
Attractive opportunities come from valuation, not growth narratives or consensus outlooks.
- Despite outside views that Japan is circling the drain and the Bank of Japan's large ETF purchases, Japanese domestic small caps offer attractive risk-adjusted returns, so Windhorse allocated to Japanese equities.
“there are pockets of opportunity to get attractive risk-adjusted returns in the Japanese stock market in general and in Japanese domestic facing small caps in particular.”
- Projected real returns on US stocks over seven to ten plus years are derisory, if not negative at reasonable inflation rates, so Windhorse does not invest in US equities broadly but finds opportunities elsewhere.
“The projected real return on U.S. stocks over seven and ten plus years is derisory, if not negative”
- Long-term capital market history shows high-growth economies with strong GDP growth and favorable demographics tend to generate subpar stock returns, so Salem's Asia ex-Japan investing relies on attractive prices and family-controlled firms rather than growth.
“high growth economies that are flattered by relatively high economic growth rates of the GDP level and by favorable demography tend to generate, surprisingly perhaps to many people, subpar returns.”
- Swensen's Yale approach aimed to exploit dispersion of results by allocating to asset classes where the best managers can transfer wealth from weaker managers, with liquidity supporting venture and buyouts.
“a wide variety of asset classes and subclasses where the very best managers could transfer wealth from other managers to themselves and their clients. That's dispersion.”