If I weren’t so committed to editorial integrity, I could’ve penned a terrifying headline for this piece. And believe me, I was tempted.
Instead, I went with a clever play on “War is hell.” That’s a good compromise: It commands reader attention without playing the fear card too aggressively.
I’ll try to keep this simple, or anyway avoid making it more complicated than it has to be.
At-scale, modern portfolio management revolves around VaR. Flawed or not, that relatively simple, probabilistic measure is the foundation of risk management for large portfolios. A corollary says everyone’s a vol-scaler. Conceptually, I mean.
Historical volatility’s a very important input in position sizing, and because it’s historical volatility, it moves higher on a lag, embedding a kind of latent sell impulse across markets during periods when realized vol increases and stays elevated.
Remember Paul Bettany’s “Will” trying to explain to Kevin Spacey’s “Sam” how his own trading operation worked in Margin Call: “As far as I can tell these here are the historical volatility limits, which I guess our whole f–king model relies on pretty heavily.”
Of course, realized vol’s still very low in US equities, certainly compared to implied. I’ve been over that more times than I can remember since the war started late last month. But when it comes to multi-asset portfolios (risk parity’s the quintessential example of vol-sensitive, levered, multi-asset portfolio management, but as noted above, pretty much everyone fits that description whether they realize it or not), realized vol’s moved up amid all the shooting.
The figure and table above are from Nomura’s Charlie McElligott. They give you a sense of where we’ve just been (i.e., what happened to realized vol across assets this month expressed as a ratio of the past decade average), where we could be going (if realized vol ebbs, stays the same or increases), and what that would mean in terms of cross-asset deleveraging flows using the bank’s risk parity portfolio replication just to get a feel for things.
There are two key takeaways. First, and as McElligott put it, “This is less about actual risk parity strategies but instead a thought exercise about any multi-asset portfolios.”
So, the point isn’t to parrot the clichéd risk parity deleveraging doomsday narrative that’s been a fixture of the financial mediasphere for over a decade now (although I certainly could’ve spun it that way, and indeed the overarching message is even more foreboding given that the thought experiment applies to the entire institutional asset management space). Rather, the point is to illustrate the directional read-across for multi-asset portfolios of this month’s increase in realized cross-asset vol under different forward-looking scenarios using a risk parity portfolio as a proxy.
The chart above gives you some context for the table included in the first set of visuals. The only scenario under which modeled risk parity gross exposure won’t be de-risked going forward is an environment where multi-asset vol halves.
“The idea here [is to] project naive, linear paths for volatility scenarios to see the impact these scenarios would have over the near- / medium-term on asset exposures using a model for risk parity,” McElligott explained.
The second key point — and most of you have probably discerned this by now — is that multi-asset portfolios could be deleveraged even in the event vol softens from here. That’s the 75% scenario.
“Volatility does not have to increase in order to generate substantial deleveraging flows because trailing rVol is still ‘averaging up’ after a sustained multi-year low-vol period,” Charlie wrote. “Even vol softening to ‘just’ 75% of ‘Iran Crisis Vol’ would drag up trailing realized off a long-horizon lookback.”
If we were to sustain that “Iran Crisis Vol” level (the 100% scenario) or, God forbid, if cross-asset realized vol were to rise further, the forward-looking deleveraging potentials are obviously much larger.
“These notional projections are simply the risk parity-specific AUM assumptions,” McElligott went on, reiterating that the exercise was conducted using Nomura’s risk parity portfolio replication.
As noted above, the nature of modern risk management means that conceptually and, crucially, directionally, this thought experiment’s relevant more or less across the board. As Charlie wrote, it “applies across much of the vast institutional asset management space, and would mean a lot of sell-flow in the weeks and months ahead as rVol ‘catches up.'”




Thanks for this, sir.
““Volatility does not have to increase in order to generate substantial deleveraging flows because trailing rVol is still ‘averaging up’ after a sustained multi-year low-vol period,” Charlie wrote”
I found this quote interesting. Sometimes when I read a comment on how “elevated” vol is in the mid-twenties, I get a twitchy thought that it’s not all that high versus some periods in the past.
I may not be following this, but those don’t seem like very large notionals. 75%, equities, 1m, sell (only) $8BN?