Of forced trades and vanishing liquidity

One of the biggest phenomena we have been paying close attention to over the past few years has been the growing preponderance of “illogical” market behaviour, arguably by equally “illogical” market actors.

Of course, the choice of the word “illogical” is ironic, since “logical” really depends on the prior belief system one assumes before making the said statement. Logic, as it turns out, is rather subjective. For most market participants, the “logic” of making trades, regardless of time horizon, is one of “buy low, sell high” (or “sell high, cover low” in the case of a short sale). The motive is to seek out an expectation of profit, and under this logic, we buy what we think will go up, and sell what we think will go down. The need to form these expectations lead to all sorts of other mental models – mean reversion, valuation multiples, trying to sell the top/buy the bottom etc.

Unfortunately, this overarching logic, while traditionally assumed to be the only determinant of trading action in the market, is certainly not singular. Over the past decade, structural changes in the makeup of participants in the market as well as the nature of financial products being offered by institutions have all come together to fundamentally change the trading motive of market participants.

We touched on one segment of this when we wrote about self-driving markets and about the gamma squeezes in Gamestop, about how the mechanics of markets and their structure are able to cause eye-watering moves to the upside in equities. These moves, as we are now starting to see, can also happen to the downside.

But our main point here is this: blow-off tops and violent sell-offs are not only the preserve of the world of equities (and crypto). As it turns out, even in the supposedly tranquil world of fixed income, and the highly giga-brained lands of derivatives and structured products, similar drivers for such involuntary behaviour exists.

Misreading them as voluntary, profit-driven actions is the first step to getting into a lot of trouble. Here are some that come to mind.

In the name of Risk Management

The discretionary manager/trader’s worst nightmare probably involves a visit from their risk manager, not because risk managers are unpleasant people, but very likely because such visits entail being forced to make trades that go against the profit-seeking belief. The only way to avoid these visits is to have the discipline to chop losing positions or put on hedges before they cause too much damage.

When push comes to shove, directives to chop risk lead to the commonly known occurrence of “flush outs” at the bottom of a cycle, where a final capitulation move from positions that are forcibly closed leads to a big final selloff before the books are cleared.

But forced trades can come in many other different forms. For example, inverse structured products that are sold against an underlying benchmark (e.g. inverse volatility) create hedging scenarios for the issuer that reflexively reinforce directionality.

Consider this: a structured note that is meant to deliver the inverse return of a volatility index (e.g. VIX) is sold to clients. If VIX goes up, the price of the note goes down. But VIX has a lower bound and no upper bound, so if a note is incepted with a price of say $10 when VIX is at 15, what happens when VIX is at 33 and the structured product cannot be priced at -$2 (no client is going to pay the issuer to own a negatively priced product). As a result, as VIX approaches the level below which the issuer can no longer hand losses over to the buyer (i.e. $0), the issuing desk has only one choice: to hedge, and the hedge involves none other than BUYING more VIX exposure, thereby reflexively pushing VIX up MORE.

Going down this line of thought, of situations where traders are made to do trades that are not necessarily motivated by a search for profit puts a lot of sense to other occurrences we see in the market like occasional bids for treasuries in a bond bear market (need for collateral outweighs trading P&L).

Whether as a function of internal risk management or as a result of regulatory requirements that limit the amount of risk banks and market makers can warehouse on their balance sheets in the name of preserving system stability, the rise of non-profit-seeking trades as a proportion of the trades that hit the market mean we need to apply additional context to the price action that we observe.

Fed pls do something

This story of being forced to do the only thing that you can doesn’t only apply to commercial banks. It also applies to central banks.

As is the case for the Fed, so it is for every other central bank around the world that, as Chair Powell mentioned in his press conference on Wednesday following this week’s FOMC meeting, there is little that monetary policy can do to impact the supply side of the economy, which is the root cause of much of the inflationary pressure we’re seeing. All the Fed (and every other central bank) can do to address the inflation situation is influence the demand side, and the only way to lower inflation from the demand side is to extinguish demand.

Whether the Fed succeeds this time in creating a “soft landing” (note: their track record isn’t great) is beside the point: what is clear is that notwithstanding the risks of recession, the political pressure to address inflation is too high.

Where in the crypto world, the communities cry out to the devs to “pls do something”, so too is political pressure calling to central banks to “do something”. And while doing something (i.e. tightening into an already tight and tightening financial environment) is arguably not the best move, the meme of “money printer go brrr” is so strongly embedded as the cause of the inflationary pain and suffering of the masses that based on political calculus, doing something is better than doing nothing.

We were taught many years ago that economics is inseparable from politics – and it turns out that on that front, our economics teachers in high school, without any PhDs or central bank experience, were right.

You give me money, I buy; You ask for money, I sell.

This is the mantra (and the trading algorithm) of the typical passive ETF, and for the past 15 years, the structural rise of passively managed funds that simply “buy the index” has been a boon for market access. Sure, it annoyed many of the traditional long-only benchmarked funds who suddenly had a good thing (fees) competed away, but for everyone else, liquidity rose, borrow costs cratered and one could almost argue that it was a net positive.

Yet to get to the conclusion of “markets always go up” as an extension of the above mantra, one important assumption needs to be made. “You give me money” must hold true for ETF managers. For most ETFs, inflows over the years have come from defined contribution pensions and 401k plans, as well as target date fund investments, and the underlying driver of ongoing contributions is ongoing employment.

One can only wonder what the impact of the Fed’s determination to engineer a slowdown in aggregate demand will be on employment, especially if they continue their track record of engineering soft landings.

But redemptions don’t necessarily only come from unemployment. They can also come from rebalances that are, once again, involuntary.

Remember the orthodoxy of the 60/40 debt/equity portfolio? That remains the orthodoxy by mandate for many pension funds, who also happen to be one of the biggest allocators to passive funds.

With the debt markets in the doldrums this year, thanks to inflation and rate hikes, and equity markets (at least the SPX) holding up relatively better (aka “down less”), one would expect the “logical” move to be to seek out a decent inflation hedge and reallocate to equities from bonds that are losing their value to inflation.

On the contrary, the 60/40 orthodoxy mandates a rebalancing to MAINTAIN that balance: at risk of oversimplifying, let’s assume a portfolio started 2022 with $1,000 notional, with $600 allocated to TLT (treasuries) and $400 to SPY (equities). As of this week, TLT is -21%, and SPY is -13%, so this blended portfolio is now worth $822, of which $474 (57.7%) is in TLT and $348 (42.3%) is in SPY.

In accordance with the 60/40 mandate, if this portfolio were to be rebalanced, it would entail therefore SELLING the outperforming asset type (SPY) and BUYING the underperforming asset (TLT). Of course, these rebalances often take place more frequently than once every 5 months, but the conclusion is the same: selling equity ETFs mean asking for money from the ETF manager, and for the algo, it’s a case of “You ask me for money, I sell”.

How many guns held to how many heads?

For many in the business, all of the above may sound like business as usual, things one would do in the normal course of business. But let’s be very honest with ourselves: these are not “normal” times.

The intricacies and plumbing of the global financial system are more complex than anything we can fully understand, much less articulate into a single post (or even multiple posts). And we’re not even talking about just internal domestic liquidity between banks, dealers and central banks – we’re talking about the global supply chain of liquidity, intertwined with the supply chain of physical commodities which, as we’ve touched on before, is itself tightening up.

The result is that many people have guns pointed to their heads (not literally, of course, hopefully…) and find themselves forced to make trading decisions out of necessity and survival.

To conclude, we’d like to end with this chart that shows the price of the next upcoming Eurodollar futures contract, plotted weekly and going back to 1982 where the time series available on TradingView ends. To us, this is the blinking signal that tells us that something out there is going awry:

For those unfamiliar with Eurodollar futures, the way to read this is as follows: Eurodollar futures are quoted at a price that is (100 - forward 3 mth LIBOR). In other words, when the price reads 97.295, it means that 3mth libor at the next contract settlement is priced at 2.705%.

Eurodollar funding represents the price of liquidity in the offshore USD market – beyond the purview of the Federal reserve. It reflects the cost of liquidity of everyone else: rising USD funding rates mean that USD is scarce in the global liquidity system.

Surprise! Wasn’t QE supposed to have printed lots of money? As it turns out, QE didn’t print money, it printed bank reserves i.e. it made it easier for banks to issue loans by providing them with more reserves, rather than the common belief of money flowing everywhere. The implications of this are deserving of an entire note in itself, but for today, all we need to know is that QE didn’t REALLY spew money all around the world. It only created bank reserves. And if the commercial banks didn’t turn those reserves into lending, that doesn’t translate into real liquidity.

And indeed, the banks are not only deciding not to lend, but they are also in fact GIVING these surplus reserves back to the Fed in the form of reverse repo transactions happening at least weekly to the tune of $1.7-1.8tn, in return for the very instruments the Fed was buying via QE: Treasuries.

Put together with the Eurodollar situation (tightening liquidity in the Eurodollar market), this makes sense once we consider that Treasuries are the most pristine of pristine USD collateral. Shortage of collateral in a tightening market = can’t borrow = need more collateral, all happening while the cost of liquidity is rising.

Ultimately, the driver of “numba go up” in markets is liquidity: when liquidity is abundant, markets can rise; but when liquidity shrinks, and the music stops, and things start to get much less pleasant on the way down than they were on the way up.

Eugene LimMay 6, 2022