Entire portfolios have been built on the adage that the world catches a cold when the US sneezes – a family of popular relative value trades allow investors to fund long volatility bets on Asian and European indexes by selling volatility on the S&P 500. But the rule was turned on its head in February’s selloff, leaving investors with an estimated $500 million in losses, and posing an awkward question: what to do when the US is the sickly one?

The answer varies. Some have chosen to limit their exposure to the short volatility leg of the trade, while others have removed it altogether – responses that paid off during a new bout of US-led panic earlier this month.

“Many hedge funds and pension funds have changed the way they see and use volatility,” says Guillaume Flamarion, head of equity derivatives group payoff structuring at JP Morgan. “The February event was a wake-up call for everyone and we saw a transition towards more defensive volatility positions, so most were better positioned for the selloff in October.”

Some of the smarter – or luckier – investors made the shift before February, when the CBOE’s Vix index of S&P 500 implied volatility jumped from 14 to 40 in just a week. For many more, the months that followed saw a scramble to restructure outstanding relative value trades such as variance spreads and dispersion, with dealers devising more defensive payouts that offer greater protection in the event US markets slide more than their counterparts overseas. 

“The sharp repricing on S&P 500 volatility has put people off what they have been doing for years, which has been to fund a long volatility position by selling S&P volatility. With a number of catalysts pointing to a possible selloff and the perception we’re getting nearer the end of the cycle, positioning has to be adjusted,” says Benjamin Corbin, head of portfolio manager solutions in the equity derivatives structuring group at BNP Paribas.

The upended dynamic is rare. The US index has been more stable than other parts of the equity market in recent years – and breaks from the norm were short-lived – tempting many firms into the trades. This time, however, the inversion has been persistent. The Vix traded above its European counterpart, Vstoxx, which measures volatility on the Euro Stoxx 50, for most of March and April. 

And October’s round of market jitters sent the US volatility index soaring from 15 to an intraday high of 28.84 in a 48-hour period – almost five points higher than Vstoxx and still quoting at a premium a week later.

The February event was a wake-up call for everyone and we saw a transition towards more defensive volatility positions

Guillaume Flamarion, JP Morgan

Volatility traders are not turning their backs on relative value vol strategies just yet. It just demands a more cautious approach.

“We expect volatility arbitrage as a strategy to remain relevant, but we feel the systematic nature of vol arbitrage in the near term may not be as attractive as before and we have to be more tactical in terms of opportunities,” says Vinod Nehra, head of the systematic investments group at Coutts, the private bank and wealth management arm of Royal Bank of Scotland. “The vanilla spread remains attractive, but there is a big risk of substantial drawdowns.”

Up and away

So, what can investors do? Broadly, responses fall into three buckets: limiting potential losses on the funding leg; scrapping the funding leg altogether; or selling volatility on an index other than the S&P.

In the first bucket, so-called “up variance” has emerged as a popular approach. A typical variance spread sees an investor sell S&P 500 volatility to fund a long position in Euro Stoxx 50 or Nikkei 225 volatility. The defensive alternative sees investors remove the extreme downside tail by limiting the level at which they sell volatility.

A pure variance swap reflects realised volatility of the underlying index and can be replicated by a strip of options across the full range of strikes.

Up variance sees investors sell a variance swap on the S&P 500 that reflects realised volatility only for returns that are above 50% of the initial spot level of the underlying index. The short leg of the trade is replicable with a strip of vanilla options for 50% strikes onwards and means investors do not sell volatility where spot is below 50% of the original level.

By omitting the extreme downside, the trade may be less lucrative in terms of carry compared to a standard trade. But proponents claim the defensive alternative would have delivered improved performance through this year’s US market selloffs and can be entered at cheaper levels compared to the standard version.

If you cut the extreme tail in a relative value regional spread, you can enter the trade one point cheaper than the full variance spread, while at the same time mitigating the downside risk

Antoine Garaïalde, JP Morgan

“If you cut the extreme tail in a relative value regional spread – Euro Stoxx versus S&P 500 – you can enter the trade one point cheaper than the full variance spread, while at the same time mitigating the downside risk,” says Antoine Garaïalde, head of payoff structuring in the equity derivatives group at JP Morgan in London. “If you look at what happened in February and October, this would have helped. It also makes the product more liquid since it is replicable with a portfolio of options without the very far out-of-the-money options, which don’t trade.”

This isn’t for everyone. Coutts’s Nehra warns that while it might make sense to limit the short position from a risk management perspective, changes to the exposure of the trade might not always be desirable. 

“We are not too comfortable with those structures as one doesn’t realise the extreme moves for which one has actually entered into the trade.”

A long-only variation on the theme, dubbed “down variance” by Societe Generale, has also become popular as a hedge against a market crash of the type witnessed in February.

The strategy enables investors to build a more targeted volatility exposure by replicating variance with a defined set of options, thereby reducing the cost of carrying a long volatility position. Options are selected only within specific strike limits that reflect a falling market. That means investors can avoid paying for unnecessary upside strikes – when they don’t need protection – or the most extreme downside levels that may be unhedgeable.

“What we see is that we don’t need to pay for volatility on the upside and removing it reduces the cost of the usual long volatility trades. This gives you a targeted exposure to volatility, only paying for volatility at the same time the underlying index is going down, while maintaining the convexity we like,” says Geoffrey Simmons, equity derivatives engineer at Societe Generale.

The bank reckons the trade reduces the cost of a long volatility position to around half that of a standard put-based structure.

Long corridors

Some investors have opted to remove the short funding leg altogether. According to BNP Paribas, trades such as corridor variance, which were once largely about the volatility spread between US and European or Asian indexes, are now being reshaped to provide long-only volatility exposure. 

Corridor variance spreads have been a popular trade in recent years, rising to more than $200 million vega notional outstanding, according to industry estimates. The classic structure sees investors combine a short position in S&P volatility with a long position in Euro Stoxx 50 or Nikkei 225 volatility, but only while the underlying moves within a certain range. The trades help dealers recycle the vega exposure they accumulate from issuing autocallables – a popular class of retail structured product – as the range on the corridor variance spread reflects the twin barriers in the retail notes.

Those trades slipped underwater during the February rout as the short leg got hit, triggering estimated paper losses of up to $375 million.

Mono-corridor variance dispenses with the cross-index spread element of the trade altogether. Instead, investors take a long volatility position, primarily on the Euro Stoxx 50, within a range that reflects autocall knockout levels. There is no selling of US volatility to fund the position. The most commonly traded structure is forward starting, where investors start receiving a return in one to two years and the trade expires in three to four years. 

By combining a forward start and a corridor… you can lower the entry point and minimise the cost of carrying a long volatility position

Benjamin Corbin, BNP Paribas

“By combining a forward start and a corridor, where the knockout on the upside mimics demand in structured products as closely as possible, you can lower the entry point and minimise the cost of carrying a long volatility position,” says Corbin at BNP Paribas.

The French dealer has sold around €20 million ($23 million) vega in the products so far this year, a fourfold increase on the amount it sold throughout the whole of 2017. 

Forward volatility agreements, which see investors buy forward-starting puts, have also gained in popularity as they allow investors to obtain long implied volatility without exposure to realised vol. The structure has largely been traded on indexes that underlie structured products, such as the Euro Stoxx 50, where implied vol is cheap compared to other indexes.

Alternative shorts

Even with more defensive structures at their disposal, not all investors have been coaxed back into selling S&P 500 volatility.

“We remain cautious about using the S&P as a funding leg, but this could change if we feel vol-of-vol becomes appropriately priced,” says Daniel Benchimol, a senior portfolio manager at Wolverine Asset Management.

The CBOE’s vol-of-vol index, VVix, which tracks the implied volatility of the Vix itself, hit an all-time high of 167 in February. When the Vix jumped again on October 11, VVix hit 125.

That suggests a more contained ‘risk off’ event than was witnessed in February. In part, that stems from a shake-out in short volatility positions held through exchange-traded notes whose end-of-day rebalancing was blamed for the violence of February’s moves.

But the spring clean in listed volatility positioning has not been mirrored in over-the-counter markets.  Benchimol says an influx of new institutional volatility players, mostly active in dispersion and short vol on the S&P 500, added to already crowded trades. Those were not generally unwound following February’s mayhem, leaving the trades as crowded as ever and raising vol-of-vol risk on the US index, he says.

We remain cautious about using the S&P as a funding leg, but this could change if we feel vol-of-vol becomes appropriately priced

Daniel Benchimol, Wolverine Asset Management

“As a consequence, other indexes and sectors started to show better risk-adjusted expected returns for relative value strategies,” says Benchimol. 

Those indexes are not always available, however.

“If you reject the short S&P leg, you can always find another index to sell. But it comes down to what you can and can’t do, either due to your mandate or how well you understand that market,” says one trader at a US house. “Can you really sell UK equity vol when we still don’t know anything about Brexit, or Nasdaq if you don’t understand how it behaves?”

BNP Paribas has traded some relative value structures using the UK’s FTSE 100 as a funding leg. While the index has some insulation from Brexit given that more than 70% of its revenues are generated overseas, price risk remains in the form of currency swings on Brexit news flow. The bank has also seen pockets of interest in Australia’s ASX 200 as a funding leg, but liquidity constraints prevent widespread use.

“Economically speaking it’s hard to find an alternative short, as typically the candidates are not scalable to an order of magnitude of the S&P 500,” says Corbin.