There’s been some decent ink spilled recently on the “dispersion trade” which has profited from the epically low level of realized correlation among stocks. If winning trades attract capital and erode the margin of safety in the process, is this exposure crowded and vulnerable to an unwind? In this short pod, I lay out a 5-part, informal framework for thinking about risk-off episodes. In the process, we consider the pricing of vol and correlation. While the spill-over risk from dispersion trades gone wrong doesn’t appear to be high, the pricing of index volatility that results from never seen before levels of implied correlation offers a uniquely attractive cost of macro insurance.
I hope you enjoy and find this useful.
[00:00:00] Hello, this is Dean Curnutt 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. In Hemingway's book The Sun Also Rises, a novel from almost 100 years ago, a character
[00:00:25] is asked how one goes bankrupt. Slowly, and then suddenly, is the concise answer. In today's highly-financialized, credit-driven world, it's a phrase that should occupy at least some portion of our philosophy on risk. We've seen jump to default scenarios all too many times.
[00:00:42] Orange County in 94, LTCM in 98, Enron in 01, Amarint in 2006, just about the entire financial system in 2008. There would be countless others, of course, but for the Fed taking its unwritten third mandate of financial stability quite seriously.
[00:01:00] I mean, with liquidity plummeting and the global dash for cash intensifying, the Treasury bond-basis trade did seem a tad vulnerable in March of 2020, no? Absent to Fed doing QE in quote, amounts as necessary, it's less than obvious that
[00:01:14] folks on the wrong side of this trade could have withstood the mark-to-market pain. The subject at hand in this short pod is the interplay between crowding, correlation, and convexity. In the process, I'll bring to life the manner in which risk on and risk off for strange bedfellows.
[00:01:30] When my 14 minutes in 2000 words come to a close, I'm hopeful that I've made Sir Hyman Minsky happy. His time as an economist preceded the heretofore-sighted epic unwinds. He passed away in 1996. And that makes his framework for contemplating financial instability all the more interesting to consider.
[00:01:49] I also hope, as usual, to give you a chuckle, or at least as George Costanza would say, a tee he. Lastly, as with all Alpha Exchange podcasts, I aim to play some small role in expanding your thinking on matters of risk and reward.
[00:02:04] On my most recent discussion, there's no crying in correlation. I both marveled at and warned of the unusually low level of correlation among stocks in the S&P 500. Over the last month, the realized correlation between Nvidia and Apple is negative 21%.
[00:02:20] As Bud Fox said to Gordon Gekko in Wall Street as Blue Star Airlines surged, I don't know what to make of it. Outside of levitating stock prices and robotic demand from index funds, the super caps share little in common these days by way of price action.
[00:02:34] Overall, the realized correlation of stocks in the S&P over the last month is 4%. So says the Bloomberg VCA page. If, as I say too often, RealizeVal rules the world, then quote, cascading correlation causes crowding. Wow, that's some appetizing alliteration.
[00:02:53] Let's back my statement up on correlation with some numbers. First a little widget I use and you can now as well. 4 to 5 correlation points is good for one vol point. Meaning, if I've got an index of stocks with a vol profile that results from a combo of
[00:03:09] the level of vol among the single stocks and the degree to which those same stocks are correlated, if I move correlation up by 4 to 5 points, it adds a vol point to the index. Over the last month, S&P RealizeVal is 8.6.
[00:03:24] If I pop the correlation up to 24%, still exceedingly low by historical standards, the vol goes to around 13, a 50% increase in realized. It ain't peanuts as restaurant owner Sonny said to Joe Pesci and good fellas. Staring us in the face are profoundly and I would argue unsustainably low levels
[00:03:44] of correlation among stocks. Sure, Apple and Nvidia operate in different businesses, but they are exposed to a substantial number of the same macro factors. And there's no rocket science involved in uncovering that the very same conditions that
[00:03:58] cause stocks to become more correlated are the same ones that cause them to become more volatile at the same time. But for now, the stock market is its own magical diversifier. The cascading correlations acting as a GLP one for risk, shedding realized volatility,
[00:04:15] trimming the waistline of daily moves in the index. Over the last two months, the average of the absolute value of daily percent moves in the S&P is 50 basis points. Skinny! Oprah Winfrey may need to meet the fine folks at Standard & Poor's.
[00:04:30] There's been discussion recently on the quote dispersion trade and the vulnerability that may arise from a forcible unwind of this risk exposure on the books at skinny levels. I like what newfound research CIO and previous Alpha Exchange guest Corey Hofstein
[00:04:45] did in a recent tweet, proclaiming quote, the dispersion trade will blow up. His plan, tuck this away for now but have it at the ready should said prediction come to pass. Talk about warehousing optionality. Brilliant. Will the dispersion trade blow up? It's unknowable, but at least worth consideration.
[00:05:04] The trade leaves an investor generally short tails and as Harley Baspin likes to say, when a market crisis event occurs, convexity is often lurking at the scene of the crime. Trades that do well in the quiet are essentially bets on sameness.
[00:05:19] Shock the risk taking system in some way and the dispersion trade which leans especially into selling index volatility as part of the trade is going to be repriced in an unwelcome way. The question, of course, is around spillover.
[00:05:32] I can recall like it was yesterday, the frenetic unwind of short dated S&P 500 variant swaps that occurred in June 2006. In two short days, the VIX popped by five to just about 24, more than a 30% increase. The S&P move on both days down around 1%.
[00:05:51] That's a lot of vol beta. The reason Amarith was suffering accelerating losses in its high flying NET gas portfolio and needed liquidity stat as they say in medicine, it was a buyer in a hurry of short dated S&P variants. We've certainly seen episodes of spillover before.
[00:06:09] LTCM is one I've covered quite a bit and was formative for me personally with respect to how violent market prices can become during an unwind. For the market in the summer of 98, LTCM's troubles and swap spreads and long dated equity
[00:06:22] ball spilled over into European share-clash trades like Royal Dutch Shell. 20 years later, the mechanical unwind of short VIX futures trades had a huge impact on the S&P 500 options market and system wide liquidity at large.
[00:06:39] In contemplating the potential for spillover, I'd like to share a framework I utilize to consider the manner in which risk off events occur. That is, what conditions bring them about and how can we understand their degrees of severity.
[00:06:52] First, and this comes from a conversation I had with Neil Chris many years ago, something occurs that leads to a shattering of market consensus. I like to call this new news. What the market believed to be turns out not to be true or at least cast in doubt.
[00:07:08] A few examples might be from the original tech bubble that the internet meant an entirely new valuation paradigm. Or in 2006, housing prices cannot fall nationally and the related 2007 subprime is contained. Or how about in Europe in 2011, the view that a developed market sovereign cannot default?
[00:07:29] When a strong consensus view is held, it inevitably gets baked into market prices. In 2021, the correlation between stock and bond prices implied by light exotic instruments like rate contingent puts was negative. And why not? The realized correlation was very negative as well.
[00:07:46] But in 2022, the market would vastly reprice this correlation dealing a giant blow to risk parity strategies in the process. Today, implied stock-to-stock correlation is as discussed incredibly low. And this pricing isn't in just short dated options.
[00:08:03] In fact, one year implied correlation is 24% in the S&P, a giant zero-width percentile. That price tells us something about how deeply held this view is. A related second aspect of our framework for thinking about market vol events is to consider the setup and pricing.
[00:08:21] When prices are especially stretched, the risk of an accident is greater. Low rates, low vol, low credit spreads, low compensation for bearing term premium risk. Of course, the signature aspect of the pre-GFC leverage buildup was levels of risk premium that melted away.
[00:08:38] We certainly saw the same in late 2017 in the precursor to the XIV Unwind. When Bernanke first introduced the concept of tapering bond purchases in May of 2013, 10-year real rates were clearing the market at an almost absurd negative 75 basis points.
[00:08:55] That left him plenty of room to be shocked higher. The starting point matters, and when market prices leave little margin of safety, repricing could be very swift. By August of 2013, those same 10-year reels were at nearly positive 50 basis points.
[00:09:10] The sell-off in the Indian Rupiah from May to August of 2013, 11%. Monetary policy spillovers are a thing as I covered on a recent podcast with Raghuram Rajan, who headed the Reserve Bank of India in the aftermath of the taper tantrum.
[00:09:27] The third factor I'd like to propose with respect to market risk events is the reaction function of investors. When the ability to tough it out is limited by some combination of leverage and exposure to market risk, prices can suffer from force selling or hedge rebalancing.
[00:09:44] The old adage that short vol traders quote eat like birds but shit like elephants is probably worth a mention here. When the market's risk profile is exceptionally short vol, a move in underlying prices can cause derivatives hedging that reinforces that move, either higher or lower.
[00:10:01] When these positions are sponsored by highly marked-to-market sensitive investors or products, you have the makings of a pile-on effect in which deteriorating prices cause losses that lead to further position on wines. See LTCM and XIV.
[00:10:16] As we explore how these market vol events materialize, it's difficult not to consider the liquidity transformation that inevitably occurs during long periods of stability. Our entire fractional banking system is based on the concept that not everyone will need their money at once.
[00:10:33] And questions around the proper level of reserves are really at the heart of our regulatory system and banking. But like pornography, one can likely identify improperly liquidity transformation when it appears. When risk premium levels get skinny enough, the next step in the financial system playbook
[00:10:49] appears to reach for yield via illiquidity risk. The Third Avenue Fund was forced to suspend redemptions in December of 2015 as it held too many illiquid credit assets even as it promised daily liquidity. The high-yield VIX, ticker VIX-HY, soared from 175 at the start of December 15 to 320 two weeks later.
[00:11:13] Muhammad El-Iran called this late cycle phenomenon the quote liquidity mirage, which I always liked. I liked it even better when Joe Pesci said to Spider and Goodfellas, what am I, a mirage? I digress.
[00:11:25] And this leads me to the final potential contributor to market risk events, and that is the presence of a systemic risk component. This is not good. It's when banks and complex financial intermediaries become a central aspect of the market risk dynamic.
[00:11:39] When counterparty uncertainty is introduced, the potential for contagion goes live. Just ask Tim Geithner, who said quote, systemic financial shocks, the shocks which involve panics and run are fundamentally different and more dangerous than other types of financial shocks.
[00:11:57] Played expertly by Billy Crudup in Too Big to Fail, Geithner is seen on the racquetball court exasperated and out of breath telling Paulson quote, there's no bottom, we need to do something now.
[00:12:10] It came and went relatively quickly, but the SVB debacle was a fresh example of how quickly a financial entity can become troubled and for that distress to impact firms who have either direct exposure to it or suffer from the erosion of confidence that results.
[00:12:26] The Geithner and now Yellen Doctrine is to use overwhelming force, throwing money at the problem first and only asking questions later. There's no doubt that a tepid response like that pursued by policymakers during the early
[00:12:40] days of the Eurozone crisis risks being called out by the market as insufficient. Precious time can be wasted in the process and the deterioration of prices becomes a bigger challenge to arrest, but doing too much and too soon creates a host of issues
[00:12:55] as well that lands somewhere in the category of moral hazard. So back to the dispersion trade, what components of the five part framework resonate as cause for concern? Well, certainly the trades are being done at skinny levels.
[00:13:09] When realized correlation is basically zero for a month, one can sell implied correlation at really, really low levels and still make money. There are certainly market to market sensitive investors in the trade, but it's unclear
[00:13:22] that there's a huge amount of copycat capital doing the exact same thing in the same way that the street copied LTCM's trades in the late 90s. Short equity correlation trades were fashionable pre-GFC, but were often done through variants
[00:13:36] and correlation swaps, both considerably more illiquid than today's construct, which appears to rely much more on listed options. There are surely some OTC risk recycling trades out there, but anecdotally I get the sense that folks have shortened the maturity profile of the trade and as mentioned
[00:13:52] relied more on listed options for implementation. That's risk reducing from my standpoint. Circling back to the potential for systemic risk, we should never close the door on this. That said, banks today are vastly different risk takers and less leverage than they were in the pre-GFC period.
[00:14:09] Their prop trading activities and dispersion do not feel like a vulnerability that could fan out and create spillover issues. With all of this in mind, the market price of insurance feels too low.
[00:14:21] To be sure, it's certainly not low in the context of one month S&P Realized Vol that's below 9. But versus the set of forward-looking uncertainties around macro factors like inflation, rates, treasury issuance, China and the US election, equity index insurance at today's very depressed
[00:14:38] percentiles is worth entertaining as you contemplate playing defense. If you're along the S&P, you're up 15% and it's not even mid-year. The cost of a 10% out of the money put to your end, 1.15%. Food for thought?
[00:14:53] As always wishing you a great week and if you are in the Northeast, stay cool. Catch you next time. You've been listening to the Alpha Exchange. If you've enjoyed the show, please do tell a friend.
[00:15:05] And before we leave, I wanted to invite you to drop us some feedback. As we aim to utilize these conversations to contribute 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 alphatexchangepodcast.com.
[00:15:23] Thanks again and catch you next time.