Research house of the year: Societe Generale

Risk Awards 2021: quant group’s tail-risk hedging strategies ‘saved the books’ of some big clients

Andrew Lapthorne, Societe Generale’s head of quantitative research

Even before Covid-19 kicked the legs from under them, buy-side firms were having an uneasy start to 2020, says Andrew Lapthorne, Societe Generale’s head of quantitative research. “Investors knew we’d had a period of economic weakness and that central bank intervention was driving polarisation in valuations. The risk that all of that could unravel was at the front of people’s minds.”

SG’s quants were uneasy too. Since 2016, they had felt markets were strained and that a chaotic unwinding of crowded positions was increasingly likely. For several years, they had been working on ways to cushion the impact of such an event.

“We just didn’t know when it would happen, and we didn’t know how,” says Sandrine Ungari, head of cross-asset quantitative research. “So we were trying to create systematic hedges that wouldn’t cost too much.”

Of course, Covid was not the sort of crisis the SG quant research team anticipated. But when the pandemic began to spread in March, the team’s tail-risk hedging ideas worked all the same. “They’ve been thinking about this for many years,” says a satisfied client.

Investors say several lines of SG’s research proved their worth in 2020, citing work on option-based tail hedges, intraday momentum and balance sheet weakness at US corporates.

During the worst of the 2020 market volatility, the bank’s proprietary ‘synthetic down VAR’ strategy was up 60% for the year. “It saved the books of some very large clients,” says Kokou Agbo-Bloua, global head of economics, cross-asset and quantitative research for SG. “Not only did it monetise the collapse in the market, but it managed not to lose all its money as the market rebounded.”

Kokou Agbo-Bloua, Societe Generale

SG’s intraday momentum strategy gained 23% in March and gained again in June. And the team’s strategy of favouring companies with strong over weak balance sheets – often called its “zombification strategy” – climbed 14% in March, when credit spreads were widest.

In November, when news of vaccines turned sentiment against stocks that had been boosted by Covid, such as Peloton and Zoom, SG’s case for holding on to some value positions looked prescient.

“Value is important for diversifying valuation risk across all the other factors,” Lapthorne had warned in October. “There are plenty of things wrong with the value factor but there are also plenty of things wrong with asset markets overall.” By the end of November, MSCI’s US value index was up 13%.

SG’s success goes beyond the performance of strategies at key points in the year, though. Clients say they find the team more approachable than some rivals. One investor talks of a regular “ping pong” of ideas. Another describes Lapthorne and his colleagues as intellectual “sparring partners”. Clients say SG is helpful in backtesting ideas, sharing and checking code for investing algorithms and running an eye over datasets to help identify hidden bias.

Lapthorne sees each client’s problem as different. “We try to prioritise being available.” 

Tail protection

SG created its synthetic down VAR strategy specifically as a way for investors to gain tail risk protection at a cheaper cost of carry.

The idea is to systematically delta hedge a portfolio of put options. But SG’s quants found that conventional delta hedging using the Black-Scholes formula leads to over-hedging in rising markets.

Black-Scholes treats volatility as constant, Ungari explains. But markets are generally less volatile when going up. So, the bank developed models to hedge less aggressively and thereby reduce drag on returns. Investors benefit more when markets are positive at the cost of reduced gains when markets fall. It’s a trade-off.

Sandrine Ungari, Societe Generale

“If you’re up 30% instead of 32% when markets crash, you don’t care,” Ungari says. “But you do care if your carry drag is 1% versus 5% a year.”

Success for the strategy is providing protection, but also catching the upside in a recovery, she says. “That’s exactly what happened last year.”

SG formulated its intraday momentum strategy to capitalise on forced buying from hedgers, ETF providers and passive investors that increasingly causes markets to trend strongly at the end of the trading day.

The strategy tends to do well when liquidity is thin, says Ungari, making for big gains in the strained markets of mid-March.

SG published research in November analysing supply and demand dynamics in limit order books that clients say sheds further light on what happened in March and on the drivers of intraday trends.

When order books are thin and investors rush to make a trade at whatever price they can get, trends dominate. When liquidity is plentiful, markets tend to mean-revert, say SG’s quants. The team is now exploring the use of signals from market tick data to time its strategy depending on conditions.

Playing to strengths

Balance sheet weakness as a theme came to the fore when Covid-hit revenues put corporate leverage in the spotlight. SG has been working on its research since 2014, originally creating a strong-versus-weak balance sheet strategy with the risk of rate hikes in mind.

The strategy gave back most of its March gains after central banks stepped in to support credit markets and after November’s rotation into more-leveraged value stocks. But clients continue to take a close interest in the research, says Lapthorne. “The problem hasn’t gone away. It’s just been delayed.”

Of course, SG as a firm has faced well-publicised difficulties this year, losing money in its structured products business during Covid and cutting roles in equity derivatives. But client views of the quant team seem unaffected.

Buy-siders appreciate SG’s openness in how it tackles problems. Several have used the team’s work on machine learning to develop their own models and say they’ve had good results. One welcomes the team’s “humble” feet-on-the-ground approach to the topic.

The stability of the team also means SG’s quants can avoid reinventing the wheel – Lapthorne has been with SG for 13 years, Ungari for 14. “They don’t need to restart afresh every couple of years,” says one client. “They can say when something is not a new idea and point to what’s been done already.”

SG’s melding of strategy and quant modelling work wins praise too. Systematic strategies must increasingly account for economic conditions, says one client. An investor’s view on equity value might be driven more by where US rates and credit are headed than whether stocks are cheap, for example. And SG’s approach meets that need.

“Anything that’s purely quantitative, I’ll do myself,” he says. “It’s more useful for analysts to take strategic views and use their models to make calls rather than just publishing model updates.”