Two years ago Ruffer’s investment chief Henry Maxey promised us a full-blown liquidity crisis as the Fed tightened monetary policy. All we got were a few gormless US banks embarrassing themselves.
In the UK asset manager’s latest annual review (out last week but we only got around to reading it) co-CIO Maxey revisited his 2022 prediction, and mused on whether the danger has “been averted, or merely postponed”. You can probably guess the conclusion:
We don’t know exactly when the next sting jet of illiquidity will hit markets but, when it does, the financial storm will be declared ‘shocking’ and ‘out of the blue’.
Indeed, Maxey reckons that the danger of a liquidity shock causing a 1987-style market meltdown is greater than before, due to some “emergent features” of the financial system that Ruffer’s co-CIO thinks will amplify any shock.
The common thread running through all these developments in financial market architecture is that they are likely to amplify short-term liquidity squeezes and price volatility. This matters because a lot of money is managed systematically these days, with risk asset exposures scaled mechanically according to volatility and price trend signals. And, if you include money which is not run systematically but is effectively governed by backward looking volatility-based risk metrics, that probably accounts for the majority of the asset management industry. So, if liquidity, correlation and volatility become an amplified feedback loop and this disrupts price trends, enormous selling flows can be unleashed. And today flows seem to matter more than fundamentals.
Let’s go through them quickly.
To help manage interest rates in the excess reserves era the Fed introduced a new tool called the overnight reverse repurchase facility , or O/N RRP.
In short, the Fed sells a bond from its portfolio to investors and simultaneously agrees to buy it back at a slender mark-up the next day. Basically, it’s like the Fed borrowing money overnight, or investors parking it with the Fed for a tiny interest rate. The amount of money stashed away in the RRP has been falling sharply as the Fed has shrunk its balance sheet and money markets have found better opportunities elsewhere.
But Maxey thinks that could abruptly reverse if investor sentiment suddenly sours, making RRP look very attractive for everyone and sucking money out of financial markets more broadly.
The rate is set five basis points above the lower bound of the Fed’s interest rate band, which will look very attractive when investors start to fear the downside in risky assets again and bid up the price of riskless T-bills.
. . . There’s a tipping point beyond which the dash to cash becomes self-reinforcing. Central banks would probably cut interest rates in the event of any serious financial stress, but the aggressive rate cuts needed are likely to be reactive, not pre-emptive.
FT Alphaville has already written extensively about the growth of multi-strategy/multi-manager hedge funds , their increasingly large market footprint , the rise of copycat strategies and how returns outside of the top players have actually been fading.
But Maxey really isn’t a fan of pod shops either, calling them a “massive moral hazard machine” that will probably be a major culprit in the next crisis. We’ll quote at length here, with Alphaville’s emphasis below:
The multi-strategy hedge fund model is becoming a victim of its own performance success. This model allocates capital across lots of independent portfolio manager ‘pods’. It then fillets out unwanted risks, leverages up and aggressively manages the capital allocation across those pods, using stop losses. Long-run historic performance for these funds, especially the blue-chip ones, has been eye-wateringly impressive.
. . . As a result, these strategies now collectively manage between $300 billion and $600 billion, which is, on average, leveraged three to five times. To keep up with the growth, there has been aggressive competition for talent and aggressive increases in fees. This includes the notorious pass-through model, which can lead to performance fees being paid to individual pods even when the overall fund has not generated a positive performance.
This has become a massive moral hazard machine. Individual portfolio managers are incentivised and expected to max out their risk budget in the hope of collecting performance fees. When strategies don’t work or traders get stopped out, they can generally find another seat at another fund without too much difficulty. The clients carry all the risk and most of the costs. It is the reincarnation of Wall Street’s proprietary trading desks in asset management, with upside-down incentive structures, without the same regulatory oversight and with ill-fated stop loss risk management.
. . . Some of the strategies employed by these players, such as dispersion trading, have become crowded. In an unwind, they could turn a low correlation, low volatility market into the inverse very fast. Something akin to this happened in August 2007 — a fire sale liquidation of quantitatively constructed portfolios which revealed a systemic risk in this part of the hedge fund industry.
Portfolio insurance — a dynamic hedging strategy based on stop losses which was designed to let pension funds hold a higher equity allocation ‘safely’— was deemed the villain of the 1987 crash. It would not surprise me if multi-strategy hedge funds were similarly vilified after the next crisis.
Maxey seems a bit more sanguine about the 0DTE phenomenon , but is instinctively nervous ( which is fair enough ) given how frenzied it has become.
My concern is less that this ecosystem is risky now, per se. It is more how it will interact with other parts of the market during a period of stress. The potential for toxic combinatorial chemistry worries me most. For example, during financial stress different players may seek to use this market to lay off risk, disrupting the market’s normal balance. Or stress could change the behaviour of existing players in the market.
Another FTAV fave . If you don’t spend at least some of your day thinking about CCPs we can’t be friends. Maxey’s concern is less whether too much risk has been pooled in clearinghouses and more that the shift away from bilateral derivatives has mechanised already-procyclical variation margins .
Since the 2008 financial crisis, there has been a push to move as much bilateral derivatives business — eg interest rate swaps — as possible onto central counterparties (CCPs) in order to reduce counterparty credit risk concerns in stressed markets. These efforts have introduced different risks. In particular, margin requirements tend to be pro-cyclical, meaning they can increase during periods of market stress. This can create a liquidity squeeze, as market participants need to sell assets quickly to meet margin calls.
A perennial classic ever since the 2010 flash crash. Of course, it’s notable both how common small, inconsequential flash crashes have become since then and the lack of any real major catastrophes. Not that Maxey is reassured.
We have already seen algorithmic market making fail under stress. It has attracted some scrutiny but remains a feature of markets. Algorithmic market making improves liquidity when markets are operating normally but detracts from liquidity in tail events.
. . . If liquidity provision disappears when markets behave unusually, you have an amplification mechanism — and one which is much faster than human specialists. In the modern age, as we saw in the case of Silicon Valley Bank, runs happen far faster than was ever imagined in the past.
The overall effect, according to Maxey, is a closely coupled financial system that has become dangerously sensitive to increasingly pro-cyclical liquidity dynamics. As a result:
People often think financial catastrophes occur because herds of humans panic when the emotional pendulum swings from greed to fear. The next market sell-off will be much more mechanical, mathematical, precise and fast. Regulators and policymakers, meanwhile, are human — their reaction times are slower, with decisions made by committees.
I’m sympathetic to this argument, having previously written ad nauseam about how algos have changed the rhythm of markets ; the dangers of multi-managers ; why shocks are becoming bigger and more sudden ; how a ‘ volatility virus ’ has infected markets and the theory that “ liquidity is the new leverage ”, etc etc.
All that said, I’m a bit less worried about most of these factors than I was three-four years ago, for reasons that probably warrant a longer separate post someday. But it’s nice to see someone playing the classics, which still warrant a listen.