What this covers

Patrick talks with Dan in mid-March 2020, with the pandemic sell-off underway. Dan argues from the historical record rather than from the current news. He builds on Mordecai Kurz's rational beliefs idea, that the range of forecasts nobody can disprove widens sharply in a crisis. He pairs it with Ben Bernanke's financial accelerator, in which uncertainty makes lenders retreat and a small shock becomes a credit crunch. Crises are defined by the high-yield spread passing 650 basis points, and episodes range from a month or two to more than a year in 2008.

The middle of the conversation is the factor evidence. Dan cites the value factor beating the market in 66% of normal months but 91% of crisis months, with similar jumps for the profitability-and-investment and size factors. He also says small-cap growth, normally a poor place to be, rides the size factor out of recessions. He is careful about momentum: the worst performers going in often rally hardest in 2008–09, but he calls acting on that a dangerous timing bet. He describes crises as paradigm shifts, so what worked going in tends not to work coming out.

The rest covers how to build screens and where the claims turn contentious. Dan blends value metrics with survival metrics to avoid both overpricing and bankruptcy, and he says ROE below the cost of capital marks a self-liquidating company. In credit, he says yield above roughly double-B levels is eaten by defaults in normal times, but crises change that. He also claims human behavior is more predictable in bear markets, and that markets recover before the news does.

Sharpest takeaway

Rasmussen argues that crises make market behavior more predictable than bull markets, so investors should rely on base rates from past crises, favor small-cap, value, and high-yield assets, and resist the panic and narratives that dominate during downturns.

  • Across crises defined by high-yield spreads above 650bps, value, small-cap, and quality factors win far more often than in normal months
  • Crisis drawdowns are driven by a liquidity shock (the financial accelerator) that reverses sharply once credit returns, favoring illiquid and small companies
  • Loud bearish voices and media narratives peak near the bottom, so data and pre-made plans should override them

The argument · threads10 threads · 68 claims
0.85

Pessimistic forecasts and bearish voices mislead investors, since markets recover before conditions improve.

11 pointscentrality 5/5
  • Following Mordecai Kurz's theory of rational beliefs, at any time there is a range of forecasts that cannot be rationally disproven, and during crises this range expands massively, so no one can rule out the worst or best outcomes.

    “Mordecai Kurz, who has a theory of rational beliefs, which is, at any given time, there are a range of forecasts for what can happen and those forecasts cannot be rationally disproven.”

  • The 25th to 75th percentile forecasts on Metaculus for coronavirus deaths range from 200,000 to over 20 million, and no one can prove the higher forecast is more accurate than the lower one.

    “if you look at the 25th to 75th percentile expectations for how many people die of coronavirus, it's from 200,000 to over 20 million. So our range of beliefs - and nobody, by the way, can prove that 20 million is a better forecast than 200,000.”

  • Looking at prior crises, the expansion of rational beliefs is a recurring pattern, and the people most vocal in the media during these periods are typically those predicting the most dire outcomes.

    “the loudest voices and the people that were most vocal, in the media especially, were the ones predicting the most dire outcomes.”

  • Quotes from near prior market bottoms, including a 1969 advisor's excess-liquidation remark, a 1986 newspaper editor's economic wipeout warning, and a 2000 comment about selling good with bad, show that the most pessimistic statements came right before stocks rallied.

    “These are all, by the way, from right around the bottom. So this is right around the time before stocks went up massively.”

  • Two common crisis narratives spread through social networks: friends who sold early and seemed prescient, and the sense that news will obviously get worse, and these narratives are typical of crises.

    “those two sort of memes that spread through networks are sort of the very common things that we see in crises”

  • Markets price in events well before they happen, so markets will rise before business conditions and the pandemic improve, and recovery begins as soon as the worst rational beliefs are ruled out.

    “markets will go up before business conditions start going up. And markets will go up before the coronavirus starts getting better.”

  • Over the long term, 100% of panics in U.S. equities have resulted in a return to the prior peak, showing markets are very resilient.

    “over the long term, 100% of panics in U.S. equities have resulted in a reaching of the prior peak, that markets are very, very resilient.”

  • The most bearish people are the most visible during crises because they were recently right, and they remain bearish, so media and social networks are saturated with pessimistic opinions near the bottom.

    “The most bearish people from three months ago, all of a sudden, get all the media attention, because they were right.”

  • The data suggest the true bottom often comes three to six months after the panic starts, so the next few months are probably the best buying opportunity in stocks and bonds in a decade.

    “the data would also suggest that the true bottom often comes three to six months after the panic starts.”

  • Investors should set aside pessimistic voices, experts, and friends, and instead rely on data, base rates, and plans made before the crisis, starting with high-yield bonds and then small-cap and value equities sized to risk tolerance.

    “now is the time to, more than ever, rely on data and rely on plans that were come up in advance of these crises, rather than relying on narrative, panic or even expert forecast”

  • The most important lesson from studying crises is to rely on data, base rates, and plans written in advance rather than narrative, panic, or expert forecasts, even when those forecasts are dire.

    “now is the time to, more than ever, rely on data and rely on plans that were come up in advance of these crises, rather than relying on narrative, panic or even expert forecast”

0.69

Illiquidity-driven selling and returning liquidity create opportunity in small, financing-dependent companies.

10 pointscentrality 4/5
  • Ben Bernanke's financial accelerator explains why small shocks produce big crises: uncertainty causes banks and investors to pull back from lending and buying, which cuts off credit to firms, lowers hiring and investment, and deteriorates real-economy conditions.

    “So the financial accelerator is Bernanke's response to something called the small shocks, big crises puzzle, which is why do small shocks lead to big crises.”

  • With high-yield yielding about 8% and double-Bs about 6%, pensions with a 7% return hurdle should buy high-yield now, since it is safer than equities with a smaller drawdown and could return high teens or low 20s if spreads normalize.

    “For pensions funds that have a 7% return hurdle, this is a good time to be buying high-yield. It's going to be safer than equities.”

  • In the last few crises, high-yield returns outpaced large-cap stock returns in the first 12 and 24 months after a recession ended.

    “in the last few crises, high-yield returns have actually outpaced large cap stock returns in the first 12 and 24 months out of a recession.”

  • When liquidity returns, the sharpest whipsaw up happens first in companies whose decisions depended on liquidity access, because a company that receives a loan hires contractors and its earnings forecasts jump, so markets recover sharply merely from restored liquidity.

    “when that liquidity starts to return to the market, the biggest whipsaw up is going to happen first in the companies whose decisions were fundamentally based on the availability of liquidity.”

  • The current credit freeze (high-yield spreads in the 700s, no new lending, and called-off M&A) disproportionately affects illiquid, small-cap, and micro-cap companies that depend on financing, so these firms are the ones pricing in the crisis most severely.

    “that liquidity withdrawal disproportionately affects illiquid and small companies that rely on financing. So in the public equity markets, small-caps, even worse, micro-caps. In the credit markets, obviously, high-yield issuers.”

  • Investors who provide liquidity to distressed companies, through debt or equity, capture the fastest recovery, which explains why small-cap, value, and high-yield work so well when bought in crises.

    “anyone who is willing to provide those companies with liquidity, whether through debt or through equity, that's what whipsaws out the fastest.”

  • Because so many assets are being thrown out purely for illiquidity, the negative momentum signal is weaker than usual in this crisis, and investors should lighten their avoidance of negative momentum and expect the most excessively sold assets to bounce hardest.

    “if there was ever a time to be a little bit lighter on the negative momentum, avoiding negative momentum, it's now, because there are so many things that are getting thrown out purely because they're illiquid.”

  • The most obvious move for equity investors in this environment is to shift down in market capitalization toward smaller companies.

    “the thing that stands out as the most obvious thing to do in this type of environment for equity investors is to go down in market cap.”

  • When lenders withdraw liquidity, the damage falls most heavily on illiquid and small companies that depend on financing, such as micro-caps and high-yield issuers, and when liquidity returns the sharpest upward whipsaw happens first in those same companies.

    “that liquidity withdrawal disproportionately affects illiquid and small companies that rely on financing.”

  • Companies like Microsoft, Google, and Facebook are not up against a wall and face a relatively low price for capital, whereas many small illiquid firms have only months of liquidity left and need equity or lending, which creates the opportunity.

    “There are a lot of small, illiquid companies who are up against a wall”

0.60

Value and small-cap factor premiums work far better during economic crises than normal periods.

7 pointscentrality 4/5
  • Buying companies with positive net income and positive operating cash flow matters little in easy-money periods but its returns roughly double during economic crises, so simple quantitative rules add much more value when the spread is wide.

    “even very simple ones, like buying companies with positive net income and positive operating cash flow - doesn't actually matter all that much during times of easy money. But during an economic crisis, the returns to simply buying companies that have positive net income and positive operating cash flow is essentially double.”

  • The High Minus Low value factor outperforms the market in 66% of normal-period months but 91% of crisis months, with Conservative Minus Aggressive winning 46% vs 74% and Small Minus Big winning 51% vs 71%.

    “value wins in 66% of the markets when we are at normal levels and 91% when we are at crisis levels.”

  • The research team spent about two years studying every economic crisis since 1970, defining a crisis as any period when high-yield spreads exceeded 650 basis points, which they treated as one standard deviation above average.

    “we spent about two years studying every economic crisis since 1970, which we defined as any time high-yield spreads went above 650bps. So we defined that as one standard deviation above average.”

  • Classic Fama-French factors are roughly 8x as predictive during economic crises, and the premium factors outperform the market much more often when high-yield spreads exceed 650bps.

    “there's a recent study that found those factors are about 8x as predictive during these times of economic crisis.”

  • In every economic crisis studied, small caps have led the rally coming out, and the 1929 example shows small caps recovered to pre-crash levels in four years versus twelve for large caps, with small caps rising over 800% from 1932 to 1936.

    “in large cap was 12 years but the answer in small cap was only four years. So in the four years from 1932 to 1936, small caps were up over 800%”

  • Based on base rates, value works about 55% of the time in good markets but about 70% to 80% of the time in bad markets coming out of them, so waiting for the next crisis is the answer to when value will work again.

    “if the base rates are a guide and value works 55% of the time in good markets, it works 70% to 80% of the time in bad markets, coming out of those bad markets.”

  • The study was focused primarily on the U.S. market, so the speaker is not prepared to make claims about international markets from that research.

    “We focused this study primarily on the U.S. I'm not as prepared to speak about international markets.”

0.49

Cheap, healthy, out-of-favor stocks beat expensive growth, so valuation is the best guide.

8 pointscentrality 3/5
  • Value investing relies mainly on multiple expansion to work, since cheap stocks re-rate upward and growth stocks fail to grow as expected and re-rate down, and this multiple convergence drives most of the historical value premium.

    “usually value investing relies on multiple expansion in order to work, that in your paper - Factors from Scratch - as you rebalance these portfolios, the cheap stocks re-value up and the growth stocks don't grow as much as people thought, so they re-value down.”

  • The two most richly valued areas of the U.S. market are large-cap growth, which is a large share of the S&P 500 and expensive relative to history, and U.S. private equity, which is equally expensive and the most levered part of the market.

    “U.S. large growth, so the big percentage of most total market indices, big percentage of the S&P 500 and it's very, very, very expensive relative to history.”

  • The ratio of the most expensive 10% of stocks to the cheapest 10% has reached levels not seen since 1999, and the absolute valuation of the cheapest stocks is at levels last seen in 2008 or 1999 before value rallies.

    “So the ratio between the most expensive and the cheapest stocks has gotten to levels we haven't seen since 1999.”

  • The best argument that value will not work is that today's growth businesses, like Microsoft, Adobe, Mastercard, and Ferrari, are so high quality that they generate huge cash flow and profit growth that justifies high multiples.

    “I think that the best argument to be made is that the quality of the growth businesses today is different.”

  • Because the catalyst for a shift in tech leadership is unknown and major economic events are always surprises, valuation is the best guide, and the Nasdaq is the asset priced for perfection while cheap, unloved stocks price in nothing.

    “the best way to prepare for surprises is to use valuations, to avoid overvalued things that price in perfection, buy things that are left for dead that price in nothing and if there's something that's priced for perfection now, it's the Nasdaq.”

  • Investing is not just a game of company analysis: Microsoft is clearly better than Exxon and Mastercard is better than Wells Fargo, but the meta-question is whether that quality is already priced in, and at extremes it is more overpriced today than ever.

    “investing is not a game of analysis, because on an analysis perspective, would I rather own Microsoft than Exxon? Would I rather own MasterCard than Wells Fargo? Of course.”

  • Many value stocks today are healthy, cash-generative, profitable, high gross profitability, high-ROE businesses trading at crazily low multiples because they are out of favor, with energy as a partial exception where bankruptcy risk is real.

    “they don't look like bankruptcy risks, other than maybe some sectors like energy. But you're seeing a lot of very healthy cash flow generative, net income positive, high gross profitability, high return on equity businesses trading at just crazily low multiples, because they're out of favor.”

  • Equity investing has two hazards, overvaluation and bankruptcy risk, and quantitative investing can be derived from simply avoiding both overpriced stocks and bankrupt ones.

    “I'd say if you think about sort of the Scylla and Charybdis of equity investing, in particular, those are overvaluation and bankruptcy risk.”

0.38

Crises reverse prior leaders, so the worst-performing assets tend to rally coming out of them.

4 pointscentrality 3/5
  • Crises often produce paradigm shifts so that what worked going into a crisis tends not to work coming out of it.

    “you often do see paradigm shifts following them. That what worked going into the crisis tends not to work going out of the crisis.”

  • Small-cap growth is usually a horrendous place to be, but coming out of recessions the size factor dominates so that both small value and small growth beat large caps.

    “Actually, both have beaten large cap. And usually small growth's a horrendous place to be, but coming out of recessions, the size factor plays such a big role.”

  • International markets have been and continue to be relatively fairly valued, if not attractive, and after a decade of underperformance a crisis could produce the paradigm shift that favors them.

    “international markets have been and continue to be relatively fairly valued, if not attractive. And after a decade of international underperformance, perhaps it's time for this crisis to bring that paradigm shift”

  • In 2008-2009 the worst momentum decile rallied the most, and generally the asset class that sold off worst coming into a crisis tends to perform best coming out, though acting on this is a dangerous timing bet.

    “As you pointed out in '08/'09, the stuff that got sold off worst, that worst decile of momentum rallied the most.”

0.38

A compressed value spread plus a recession catalyst should trigger a value rebound within a year.

4 pointscentrality 3/5
  • Prolonged value underperformance and widening spreads have compressed the value spring so far that it either breaks or produces a massive bounce back, and this coiled-spring setup is the basis for expecting a sharp value rally.

    “what that does have, however, though is like a coiled spring that you push down on and push down on and push down on and push down. Either the spring breaks or you see a massive bounce back.”

  • Investors have heard the value-will-work-again story in 2010, 2011, 2012, 2018, and 2019 without it paying off, which generated investor skepticism about the value factor.

    “I heard it in 2010, I heard it in 2011 and I heard it in 2012. Now I'm hearing it anew from you in 2018 and '19.”

  • The catalyst for the value paradigm shift is a recession or economic crisis, which is what the market is currently experiencing, so the bet on value will be tested within about a year.

    “To which I say, 'A recession or some times an economic crisis,' and that's what we're in right now. So maybe my bluff will be called relatively shortly, in a year from now.”

  • The speaker acknowledges he could be wrong about value recovering within a year, comparing the risk to Nate Silver's 2016 election forecast.

    “I'm the Nate Silver of the 2016 election and I'm the terrible base rate predictor of what's going to happen coming out of this crisis.”

0.30

Reaching for yield in quality credit and companies pays off during crises.

6 pointscentrality 2/5
  • In good times, easy money funds many weak ideas and bails out low-quality companies that look like bad valuations to quants, while profitable companies that fail to match investor enthusiasm can see multiples contract for years.

    “So during good times; money is easy and, as a result, a lot of stupid ideas get funded.”

  • The high-yield spread is a good contemporaneous economic indicator, and Ben Bernanke considers it the best contemporaneous indicator available.

    “We like the high-yield spread. It's a contemporaneous, very good economic indicator. Ben Bernanke thinks it's the best contemporaneous economic indicator there is.”

  • Unlike in normal markets, in a crisis reaching for yield in credit actually pays off, so investors who normally stay in investment grade should cut into high-yield now and high-yield investors should consider single-Bs, while staying selective.

    “In this type of market, however, yield is actually a little bit more beneficial. You actually tend to do a little bit better reaching for yield.”

  • In normal times, the yield-to-total-return tangency point in credit sits around double-B bonds at roughly 4.5% yield, and any yield above that is eaten up by bankruptcy losses, so higher-yielding bonds like 8% or 10% return only about 4%.

    “we basically made the argument that above that four and a half percent yield, any incremental yield you were promised wasn't going to deliver because, what tends to happen, is those companies go bankrupt at such a high rate that your yield is eaten up even more so by losses.”

  • When buying high-yield bonds in a crisis, investors should prefer companies with high free cash flow relative to debt and high ROE, and among equal yields should choose bigger companies, since size of the company, not the debt, makes bonds safer.

    “make sure you're buying companies with very high free cash flow and high free cash flow relative to debt with a high ROE. And the opposite actually of stocks; bonds, you want to buy size”

  • If GDP growth and interest rates stay at zero, investors should prize the few high-quality companies that can navigate the environment, rather than asset-heavy industrials, energy stocks, and banks that make up much of the value benchmark.

    “If, indeed, we are in this world where GDP growth is zero, where there's no economic growth, interest rates are zero, we should really prize those few high-quality companies that can navigate this market.”

0.30

Multi-factor rules blending value, survival, and momentum avoid overvaluation and bankruptcy.

6 pointscentrality 2/5
  • A robust multi-factor model should combine value metrics (price-to-book, P/E, and free cash flow yield) to avoid overvaluation, with about half weight on value, and survival metrics (operating cash flow, profitability, ROA versus cost of capital, debt-to-assets, and short interest) to avoid bankruptcy, so that no single metric dominates.

    “starting with value, let's say that's half of your model, you're going to say, 'Well, I want to look at companies that are cheap and I want to blend a variety of cheapness metrics.'”

  • Forecasters expect the decade-long trend of quality and growth dominance over value to reverse after the crisis, but only a clear recession catalyst has historically produced that reversal.

    “I think that my view is that investing is not a game of analysis”

  • A company whose return on equity or return on assets is below its cost of capital is a self-liquidating venture that will eventually go bankrupt, and quality screens should exclude such companies.

    “is this company a melting ice cube where the return on equity or return on assets is lower than the cost of capital for the company, so it's just a self-liquidating venture? You don't want to do that. That's going to go bankrupt eventually.”

  • Momentum is a signal that other cheapness metrics may be wrong: an apparently cheap stock that has fallen 90% likely has sellers who know something, so momentum acts as a risk mitigation tool.

    “we think about it as the probability that all your other things are wrong. Momentum tells you, 'Gee, this looks astonishingly cheap, but it's 90% cheaper than it was last year.'”

  • A single cheapness metric can mislead because a company may look cheap on price-to-book while holding unproductive assets, or look cheap on earnings while its cash flow is weak, so blending several value metrics is needed to identify genuine cheapness.

    “a company could be cheap on price-to-book, but they've got a lot of assets that don't generate cash flow or don't generate profits.”

  • Quantitative investors are people who have written the two lessons of avoiding overvaluation and bankruptcy into rules, and multi-factor approaches are designed to avoid both and profit by buying and holding through long-term equity and credit cycles.

    “Quantitative investors are just people who have taken those two lessons to heart and then written them into laws.”

0.25

Crisis stress concentrates in the riskiest credit, and historical drawdowns last months to years.

4 pointscentrality 2/5
  • Asset prices swing far more than real activity, as in 2009 when U.S. GDP fell 2.8% while the stock market fell 60%, and robert Shiller's question about why markets are so excessively volatile is the puzzle the financial accelerator addresses.

    “U.S. GDP growth was down 2.8% in 2009, the stock market was down 60%.”

  • Since the end of January, CCC-rated high-yield bonds fell about 15%, single-B about 10%, double-B about 7%, and investment grade fell 1% to 5%, even as rates dropped, showing the stress is concentrated in the riskiest credit.

    “Triple Cs, which are the worst part of the market for high-yield, are down 15% since the end of January. Single B is down 10%. Double B is down 7%.”

  • Crises measured by high-yield spreads above 650bps lasted from one to three months at the shortest to over 12 months in 2008, and the 2000 recession drawdown lasted about 18 months.

    “the shortest they've been there is one-to-three months and the longest, which in recent memory is 'O8, they were there for over 12 months.”

  • Historically, sell-offs of this type have produced S&P 500 drawdowns of 20% or more, and if this resembles 2008 the market could fall another 30% to 40% from current levels, though the bottom is unknowable.

    “you've tended to see losses that range, I'd say, 20% plus drawdown in the S&P 500. From here, if you look at 1987, maybe this is the bottom. If you look at 2008, it could go down another 30% or 40%.”

0.14

Crisis investor psychology is more predictable than bull markets, whose winning themes rotate by era.

5 pointscentrality 1/5
  • Investor reactions to the sell-off range from near-total panic and comparisons to 1929 to people eager to buy now, and this spread of reactions is typical of crisis periods.

    “you meet a range of different reactions from people from, 'Can you please show me the comps to 1929' to 'Should I start buying now, or should I wait a few days?'”

  • The current sell-off is one of the fastest-moving ever, with the S&P 500 nearly 30% down, small caps down nearly 40%, and international small value in the mid-30s percent drawdown, making it probably the worst drawdown since 2008 in the data.

    “This has been a very fast-moving sell-off, one of the fastest moving sell-offs ever and it's been very dramatic.”

  • The team read newspapers from each crisis to capture what people were saying, because the psychology of crises is nearly as interesting as the data.

    “go through and read the newspapers during each one of these crises to see what people were saying, because I think that the psychology of crises is just about as interesting as the data.”

  • Human psychology becomes more predictable during crises, so returns and data in crises are more similar and predictable than in bull markets, where each bull market is different, as in Tolstoy's observation that bear markets are all alike.

    “bull and bear markets are like that, too. Bear markets are all alike, but every bull market is different in its own way.”

  • Winning bull-market themes rotate by era, with SaaS software working in the 2010s, commodities and emerging markets in the 2000s, and anything internet-related in the 1990s, which is why bull markets are hard to predict.

    “what worked in 2010s was buying SaaS software. What worked in the 2000s: commodities and emerging markets. In the 1990s: anything with internet in it.”