Archegos collapse shows what can happen when leverage is misapplied

Now that the considerable cloud of dust has settled following the collapse of Bill Hwang’s family office, Archegos Capital Management, one is able to assess the damage inflicted.

The most significant loss was that felt by Archegos, which allegedly had USD20 billion in liquid assets. But with Nomura reporting losses of USD2.9 billion last week, the true extent of bank losses has exceeded USD10 billion, with Credit Suisse taking a USD5 billion (approx.) hit to its balance sheet.

Morgan Stanley and UBS also incurred losses, while Goldman Sachs and Deutsche Bank were seemingly able to extricate themselves before the fire sale ensued.

At the heart of this episode is leverage or ‘gearing’, which one typically refers to when hedge funds receive margin finance from their prime brokers to increase the magnitude of their returns; or losses. The more leverage provided to the manager, the more the PB is able to generate in fees.

But Archegos was not a hedge fund. It was a single family office, run by one individual: albeit one that had assets in excess of your average, mid-sized hedge fund.

Which opens up the question: Why did six different PBs extend margin finance to a family office, seemingly on the same underlying collateral?

Extending leverage is predicated on understanding the various risks associated with a trading strategy, and how the strategy is being executed. The more diverse the portfolio, the more comfortable a PB will be.

However, the Archegos portfolio appeared to be highly concentrated, and effectively long. It was both directional and undiversified: two potential red flags for any PB, surely? Not to say that no leverage ought to have been applied. Rather that the amount of leverage was perhaps too large, relative to the size of risk.

As Bloomberg reported, Hwang’s portfolio grew to USD100 billion. This would equate to 5x leverage in a cash portfolio. Under US regulations, PBs are allowed to extend credit on a cash portfolio up to 6.6x; i.e. a 15 per cent portfolio margin requirement.

For synthetic trades, however, such as the total return swaps used by Archegos, the amount of margin one needs to post is less, and overall leverage rises in kind. In the UK, for example, funding a contract for difference (CFD) only requires 5 per cent margin.

The point to this is that even on a very well diversified cash portfolio of long and short positions, PBs would ordinarily consider 5 or 6x leverage to be at the upper limit. So agreeing to extend leverage synthetically, on a concentrated book, does, on the surface, appear odd.

As one US prime broker expounded to me: “We focus on three things: diversity, liquidity and balance in the portfolio.”

The stocks held by Archegos – including ViacomCBS and Discovery Inc. – were unquestionably liquid. But was there sufficient diversity or balance?

Leverage is great when things are going well but it is unforgiving when the markets move against you.

Normally when a hedge fund manager presents a PB with a portfolio to finance and places collateral, the bank will enquire as to what they’re holding elsewhere: a manager can’t have the same collateral placed with multiple PBs.

In the Archegos collapse, the fact that six different PBs had exposure to the family office highlights the lack of transparency in the swap market. In a total return swap, the stock positions are held in the bank’s name before they write the swap and sell it to a client. This gives the holder of the TRS anonymity and a way to synthetically trade stocks without physically owning them.

A bank’s risk management team, though, can infer the presence of TRS by looking at where the stock ownership lies. If a rival bank is holding large positions in the same suite of stocks, it could suggest that they are sitting on the other side of a TRS trade.

It is a bit like inferring the presence of a black hole by studying the extent to which a star’s orbit resembles that of a rosette (rather than an ellipse); something known as “Schwarzschild precession”.

Both Goldman Sachs and Morgan Stanley moved swiftly to offload their stocks and reduce the impact of the resulting fallout, leaving others to pick up the pieces. Morgan Stanley apparently sold USD5 billion in shares on 25 March before Archegos hit the rails, while Goldman Sachs offloaded USD10.5 billion on 26 March.   

Neither would have wanted to find themselves in this position but it has arguably further strengthened the market dominance of the industry’s two pre-eminent US prime brokers. Not that Morgan Stanley emerged completely unscathed, reportedly losing nearly USD1 billion according to Reuters.

What happens now is likely to be a period of soul searching and a recalibration of how much leverage PBs are willing to extend. Risk parameters will doubtless be appraised, margin limits reigned in, swap agreements scaled back.

On the regulatory side, at the very least there will be a heightened level of enquiry and investigation into the counterparty disclosures regarding swap agreements.

Who is really the ultimate beneficiary of these TRS positions?

The anonymity that comes with holding swaps has been brought into sharp focus and could well lead to calls for greater transparency by global regulators. Some believe it could also lead to the curbing of investment bank activity in derivative markets.

With so much liquidity in the market, the Archegos incident is a stark reminder of how quickly things can turn when leverage is misapplied.