The Three Types of Risk-Off

Hello, this is Dean Kurnutt and welcome to the Alpha Exchange, where we explore topics in financial markets associated with managing risk, generating return, and the deployment of capital in the alternative investment industry. Hello, Alpha Exchangers, and wishing you a very happy holiday. Our great country turns 250 at a time of profound disruption when the pace of change remarkably seems only to be accelerating. If you think a little bit like I do, and I suspect you do because you are listening, markets couldn't be more interesting right now.

We are exposed to both great opportunities and threats. I've taken to saying that the two greatest assets in our country are our stock market and our bond market. Never before have Americans enjoyed so much stock market wealth. In the textbook, a recession causes the equity market to decline.

I think today's causality is reversed. Kevin Warsh ought to be watching for a sell-off in the S&P 500 for its potential impact on GDP. The wealth effect is real. This formidable stock market engine doesn't run without the bond market.

That's the other incredibly valuable asset we have in this country. an enormous, liquid, and generally risk-free government bond market. It fuels the stock market engine. With this in mind, I wanted to share with you some of the work I've recently done as part of an Alpha Exchange white paper, The Three Types of Risk Off, which I recently made public on Twitter.

Over the next 20 minutes, I'll give you the shorter version. The actual piece has some good charts in it that I hope will add to your understanding. Let's get into it. I've been in the markets now for 35 years, and there are two threads that run through pretty much everything I've explored over the course of my career.

First, what causes significant risk-off events? Do they materialize under common circumstances? Can they be anticipated to any degree? Large drawdowns wreck the process of compounding capital.

They also force suboptimal behavior. They cause investors to de-risk at exactly the wrong time. Understanding the how and why of risk-offs has been a keen focus of mine. Second and related, the study of risk premium and the price of optionality.

Why does insurance cost what it does? And how do we explain risk premium levels across assets at a given point in time? And when is it a good deal to be a buyer or a seller of insurance? Today, I want to share how I think about episodes of risk-off with particular attention to the interaction between stock and bond prices before, during, and after market vol events.

Flight to safety is a term that's been around for decades. It describes how investors retreat from riskier exposures into safer ones when volatility hits. But it wasn't until the global financial crisis that, quote, risk on, risk off, emerged as actual market jargon. It was a useful, almost necessary shorthand for characterizing the violent, fast-moving conditions of that period.

In this framework, you have a category of assets, stocks, credit, mortgages, commodities, short volatility, long carry, that benefit when uncertainty falls, but sell off together when risk perceptions rise. On the other side of the ledger, risk-off assets, treasuries, the dollar, the Japanese yen, gold. These haven assets are where capital runs when investors are unwinding risk exposures. At the center of risk on, risk off is the relationship between the U.

S. equity market and the U.S. government bond market.

Two things are happening. First, capital that runs from an asset class needs a new home, and the liquidity and, quote, risk-free nature of treasuries have made it the most compelling destination. Second, as capital crowds into bonds, investors are re-underwriting future growth and inflation expectations, incorporating the same weakness that investors are reacting to. Bond prices traditionally rally as the forward-looking outlook deteriorates.

There's also an active process of investors' front-running, anticipated Fed easing. To bring 2008 to life, looking at the 10 worst days for the S&P 500 that year, the average move higher in the TLT, our proxy for the long end of the bond market, was positive 1.8%. That's about as clean a risk-on, risk-off relationship as you'll find.

Risk-off comes in many flavors. War, inflation, debt, monetary policy, derivatives, natural disasters, elections. One way to think about any risk event, It results from the shattering of a strongly held consensus that had been baked in to market prices. As Keynes purportedly said, The setup matters too.

When the fallout lands on mark-to-market sensitive investors, de-risking happens fast and takes on a life of its own. Nothing compares to the GFC here. Years of mispriced mortgage credit risk built up tremendous leverage on balance sheets that needed overnight funding every single day. I developed a framework for classifying risk episodes according to how stock and bond prices interact There are three types classic taper and liquidation In the classic risk off equities plunge and bonds rally a flight to safety.

Up until 2022, this was the most common by far. In the taper, the bond market is the motivating factor. A sell-off in government bonds causes equities to suffer from uncertainty about how high rates will go. 2022 is the strong recent example.

In liquidation, you start with a classic risk-off, but it fails to self-correct. It runs amok, ensnaring every asset except cash and long volatility. The risk-free asset rallies first on the flight to safety, but ultimately buckles as well as investors are forced to convert everything they can into cash. March 2020 was a dangerous liquidation event.

All three risk-offs produce big increases in market volatility. What differentiates them is the correlation between stock and bond prices. In the classic, that correlation is negative. In the taper, it's positive.

In the liquidation, it starts negative and then flips positive. There's an old saying that Dr. Copper has a PhD in economics. I've taken the saying that Dr.

Correlation has a PhD in finance. There's a lot to learn from how asset prices interact with each other, for both investors and for policymakers trying to understand the speed with which they may need to act. Over the past three decades, the classic risk-off has been the most common by far. Equity sell-off, treasuries rally.

A shock arrives, investors reprice growth, inflation, profits, and default expectations, vol spikes, and capital rushes to the perceived safety of risk-free bonds, bonds that benefit further because real rates and break-evens typically fall in these episodes as well. The shocks themselves can be economic, monetary, financial, or geopolitical, and these often overlap. LTCM in 1998, the tech and telecom bust of 2002, the GFC of 2008, the Sovereign Crisis of 2011. These are the largest classic risk-offs.

The 87 crash actually started as a taper risk-off that triggered a classic one, and 2020 featured both classic and liquidation elements. The GFC is the most pronounced example by duration intensity of the classic risk-off. From September to year-end 2008, the S&P fell 29%, and the 10-year yield dropped by 175 basis points. The VIX surpassed 80.

From September to year-end, the TLT delivered a total return of 29%. The bond market did a fantastic job defraying losses elsewhere, even as nearly every other asset class suffered. If you believe inefficient markets, as I do, there are no truly positive carry hedges in this world. You wouldn't expect a free insurance policy from GEICO.

Don't expect one from Goldman Sachs either. But the bond market rallied for such a long, consistent stretch, while paying a coupon no less, that many investors mistook it for actual insurance. From 2009 to 2019, the S&P and TLT delivered annualized returns of 13.5% and 7.

3% respectively, with daily correlation around negative 45%. That's the friendly backdrop that made the 60-40 portfolio and strategies like risk parity built on top of it so durable and so vulnerable when conditions changed. In the classic risk-off, bond moves are the outcome. In the taper, they're the starting point.

Higher rates and higher rate volatility send stocks lower. The correlation flips positive. I put three events in this category, the 87 crash, the 2013 taper tantrum, and the 2022 joint drawdown. 2013 is the textbook case.

In May, Fed Chair Bernanke first suggested the Fed would begin tapering its bond purchases. His actual words, quote, That is all it took. Years of QE had pushed real 10-year yields to negative 65 basis points, a genuinely confounding price, a sign of just how price insensitive the Fed's own buying had become. The reaction to Bernanke's comments was swift.

Real yields surged 150 basis points in just four months, one of the fastest moves on record. The VIX moved from 14 to 21. the S&P drew down 5.8% and emerging market currencies like the Mexican peso and the rupiah were hit hard.

The carry trade the Fed was threatening to unwind had spillover effects around the entire world. Both 2013 and a follow-on episode in Q4 of 2018 were ultimately remedied once the Fed pivoted dovish, helped along by cooperative inflation data that gave the Fed room to maneuver. In both cases, year-over-year PCE sat around 1.6%, comfortably below target, which gave the Fed cover to ease off without losing credibility.

2022 was different. There was no air cover By June of that year CPI was running at 9 year and the Fed had to tighten into it The S fell 19 and the TLT lost 33 in 2022 Once tightening began in earnest, stock bond correlation flipped positive, 22% for the rest of the year. Joint down days, both the S&P and TLT moving 1% or more lower on the same day, had almost never happened in the prior era. Suddenly, they were common.

Here's the deeper issue. Wrong-way correlation assumptions are baked into the setup of every taper risk-off. Because stocks and bonds had been so reliably negative correlated beforehand, investors built that assumption into portfolio construction. Strategies like risk parity lever up because they expect that offset and the resulting low portfolio volatility.

Get the correlation assumption wrong and you've badly underestimated your portfolio risk. As the old Graham and Dodd line goes, there are no bad securities, only bad prices. The modern version, there are no bad securities, only bad correlations. Going into 2022, both stocks and bonds were near the 100th percentile of valuation, joint richness, not an offsetting relationship.

With 10-year yields at negative 100 basis points at the end of 2021, stocks only look cheap relative to bonds. I flagged this on the Grants podcast in December 2021, that five-year real rates implying almost negative 2% was a price wildly out of bounds and that duration's role as a risk mitigator could flip to risk accelerant. And then there's the 87 crash, the original most violent taper tantrum. These were simpler times, before the dense web of counterparty exposure, before credit derivatives, well before central banks themselves became the biggest investor in the market.

The old adage was, just as the party's getting good, the Fed pulls away the punch bowl. That year, CPI went from 1% to 4.4% over the course of the year. The 10-year peaked at 10.

25% in October, a level that would make almost anyone think twice about owning stocks. Correlation between stocks and bonds turned sharply positive in the weeks before the crash, and then flipped negative through the crash itself as the Fed aggressively eased. The move in euro-dollar futures around the crash was wild. From early September to early November, the front month contract was essentially unchanged on net, even as rates whipsawed from 7.

5% up to 9.5% and then back down again. If there is one investor credited with best understanding the flight to safety that would follow, it was Paul Tudor Jones. His chief strategist, Peter Borish, put it this way, quote, I think that Paul's greatest skill was realizing that people were going to drive to the place of most liquidity, and that was going to be fixed income.

The irony is everyone thinks Paul made all this money being short stocks, but there wasn't that much liquidity in stock index futures. A lot of the opportunity to be made was in bond futures, on the assumption that the Fed would supply a lot of liquidity. I've said it this way, equities are short the straddle on rates. In 2022, the S&P lost 19% as yields rose 220 basis points.

In 2008, the S&P lost 38% as yields fell 185 basis points. Move fast in either direction and equities suffer. The last of the three and by far most dangerous of our three risk-offs is liquidation. An extremely rare deflationary event where the value of every financial asset plunges relative to the dollar because cash itself is in such high demand that not even the risk-free complex is spared.

March 2020 is the case study, maybe the most chaotic seven days of trading in market history. Over those seven days, the S&P fell 12% and the 10-year yield rose by 60 basis points. It started as a classic risk-off. From February 19th to March 9th, the S&P lost 18.

9%, and the TLT gained 17.7%. But then the sudden stop became so severe that the rush for actual cash ensnared treasuries as well. From March 9th to March 18th, the TLT fell 15%, while the S&P fell 13% together.

Think about what that does to a risk parity portfolio. The delta loss, the correlation sign flip from risk mitigation to risk acceleration, and a surge in volatility of both assets simultaneously. Models never priced for that combination. In just 11 days, the TLT had six daily moves greater than 5%.

The dollar rallied 6% in nine days as everything. Gold down 10%, crude down 27.8%, got sold for cash. The VIX hit an all-time high of 83%.

In a liquidation, long vol is the only truly anti-fragile asset, the one thing that doesn't just survive the shock, but actually strengthens because of it. Move fast and break things is the old Zuckerberg line, Silicon Valley's hard-charging ethos that treats failure as the natural consequence of innovation. In finance fragility reorders that catchphrase to move fast and things break March 12 2020 produced the most outrageous trifecta of daily moves in modern S history down 9 up 9 and down 12%.

Those three days alone generated enough realized volatility to equal an entire year at 18.4%, even assuming the market sat closed and flat for the other 249 trading days. That's roughly three times the realized vol of the entire calendar year of 2017. The fallout hit firms like Malachite, Alberta Investment Management, and Allianz, all short convexity in one form or another.

VIX calls, variance caps, vol skew. On the other side were firms like Ionic, Longtail Alpha, Saba, and One River, and of course, Bill Ackman's now famous credit derivatives hedge, which netted a cool $2.6 billion during that window, what some have called the second greatest trade ever. The Fed had little choice but to step in.

On March 23rd of 2020, the FOMC stated that it would, quote, continue to purchase treasury securities and agency mortgage-backed securities in the amounts needed to support smooth market functioning. Jay Powell's quote, in the amounts needed, was this generation's version of Mario Draghi's 2012 Whatever It Takes. And because long vol is the one asset that thrives in liquidation, it inevitably winds up in the policymakers' crosshairs. Paul Tudor Jones understood this in 1987.

Trade the policy response. Anticipate what the central bank has to do and position for it. Two more recent episodes are worth a mention. April 2025, what Corey Hofstein calls the tariff tantrum, was a short, sharp, and self-inflicted episode.

VIX broke 50, real yields spiked 50 basis points in five trading days, and notably, both the dollar and the bond market sold off rather than rallying as a safe haven. It wouldn't have ended well but for Trump's nevermind tweet on April 9th. And more recently, we've seen the 2026 oil shock tied to the conflict in Iran. Correlations shifted meaningfully.

Oil has been increasingly negatively correlated to stocks and high yield and positively correlated to the dollar and rate volatility. Oil has acted like the VIX. I want to leave you with one more idea, a fourth type of risk-off that I think deserves real attention, a liquidation episode that starts in the Treasury market itself. This is different from the taper.

The taper is about Fed policy. It leads to higher rates and lower stocks, but the credibility of the system of government debt stays intact. This fourth category is a crisis of confidence, a sharp decline in the willingness to hold U.S.

government debt, driven by a belief that the debt load is too large, governance is too weak, and agency costs are too intractable. Here, it's the 10-year note itself that is the risk asset. If it's the wrong price, then everything linked to it gets repriced and not in a good way. It's a low-probability, high-impact tail event.

But with a trillion dollars a year in interest costs and a political system unable to touch the third rail of entitlement reform, it's the kind of risk-off I think the market is underpricing. I think about the seriousness of any risk event as scope times probability. A small increase in the odds of an extremely impactful event produces a large increase in overall risk. And I think about probability itself as difficulty times urgency.

Some problems are urgent but easy. I need to send an email in five minutes. It's urgent, but you can solve it very quickly. Others are difficult, but obviously not urgent.

Retirement planning. The U.S. fiscal problem sits in that second bucket, except the scope is global.

I covered this in depth in episode 249 of the Alpha Exchange, The Shock Heard Round the World, U.S. Government Bonds, with four guests unpacking how the U.S.

, Historically, the exporter of stability and importer of capital may increasingly be a source of risk for the rest of the world. Fed Chair Warsh said something on this podcast back in 2021. He said, if you've seen one financial crisis, you've seen one financial crisis. I believe there are real commonalities across these episodes of disruption, but each one is genuinely unique.

As Heraclitus once said, the only constant in life is change. Technology, market structure, the regulatory landscape, the global payment system, the goals of monetary policy, the geopolitical order. All of it shifts all of the time. My hope is that thinking about risk-off through this lens, classic, taper, liquidation, and maybe even a fourth, treasury-led category, gives you a sharper framework for understanding markets when things get difficult.

Thank you for listening, and I hope you enjoy the holiday. tribute to the investment community's understanding of risk, your input is valuable and provides direction on where we should focus. Please email us at feedback at alphaexchangepodcast.com.

Thanks again and catch you next time.