How can you prepare for volatility spikes by hedging feedback loops?
For the past two decades, significant volatility spikes have caused market sell-offs. How can you account for them in advance? Aymeric Kalife, CEO, iDigital Partners and Adjunct Professor, Paris Dauphine University, shares his insights into volatility spikes experienced during the May 2006, May 2010, August 2011, August 2015, January and June 2016, February 2018 market sell-offs, and how quant finance can prepare for these events.
The rally in the US market from the March crash lows has been extremely steep (+70%), with a summer rally for the Nasdaq and the US tech, followed by a sudden sell-off in the tech sector and crashing end on September 3, 2020 (-10% and -$1.9tn in just three trading days).
Such a melt-up in many big technology stocks has been fuelled by undue optimism initiated by Softbank – dubbed as the Nasdaq whale (a heavy hitter with the power to move markets on his own) – exacerbated by a wave of small-sized retail investors who poured most of their savings into stocks and call options.
Never before has Wall Street perhaps noticed such heavy demands in call options. Call options volumes started spiking in March, April, and May as retail investors opened up Robin Hood accounts and started a frenzy of day trading for smaller and smaller contract sizes. In June, SoftBank, which had previously bought $4bn worth of stakes in 26 of the world’s largest technology companies ($1.04bn of Amazon, $475m of Alphabet, $250m of Adobe, $189m of Netflix), began trading collar option trades. During August, Softbank bought another $4bn worth of 3-month and 6-month maturity listed call spread options on 6-7 big US technology stocks with a notional value of about $30bn, while momentum-oriented retail crowded small sized traders (lots of 10 contracts or less) together spent almost $40bn OTC call option trades just over a month, using one-week or even one-day options (in contrast to institutional investors that tend to favour one- to three-month options).
The one-sided massive rally in big US Tech companies happened not because they were “fundamentally great companies”, but rather as the result of aggressive bets placed by those investors who levered up gigantic positions in those US stocks through call options thus altering the price arithmetically by impacting the supply and demand of shares. For example, Amazon’s call volume averaged 146,000 in the 30 days through Wednesday, nearly a record, over a stretch when the stock jumped 9%; Apple Inc. calls averaged more than 3 million per day, the most in six years, while the stock rallied 24%; Tesla Inc.’s daily call volume headed toward 2 million while the shares climbed 28%.
The combination of
Actually, the short-term and close to at-the-money nature of the contracts required hedging by market makers who in turn triggered a feedback loop due to their massive amount, whereby market makers bought massive stocks and VIX contracts to hedge their massive short net positions in call options, especially for short term options and if it starts moving towards the strike price. This boosted further higher US Tech stocks and fulfilled the investors’ own upside prophecy, while contributing to the curious phenomenon of the S&P 500 Index rising at the same time as VIX Index. Reversely, when stocks dipped in early September dealers instead ditched their hedges, exacerbating the sell-off. Retail didn’t have the ability to move the market by themselves, but by buying calls they forced dealers to hedge themselves, and triggered this parabolic move in tech stocks.
Over the past decade, there were more short-lived but sharp transitions from low volatility to high volatility with no well-known fundamental catalysts than in the prior two decades, which is illustrative of a new market regime and illustrative of the growing market impact of hedging by increasingly larger derivatives dealers. Although fears about growth or sovereign debt sustainability may have contributed to the significant volatility spikes experienced during the May 2006, May 2010, August 2011, August 2015, January and June 2016, February 2018 market sell-offs, they do not fully explain either the extreme magnitude of the shocks or the repeated occurrence.
Actually hedging feedback loops by large players have exacerbated the acuity of volatility from the illiquidity in option markets since the late 1990s, stemming from a structural imbalance between supply and demand in derivatives, as highlighted by a former Head of Research of the NY Fed, John Kambhu. This is notably illustrated by the growing hedging needs of U.S. banking mortgages and insurance annuities, and Asian structured products and the lack of sufficient natural counterparts to meet their demand, which only large dealers can meet by selling puts and calls. These massive imbalances in the derivatives markets is at the source of hedging inefficiencies translating into hedging feedback loops, as large dealers makers need to hedge their massive net short derivatives positions, which requires buying (selling) the underlying asset after its price rises (falls). rises, in transaction size that are large enough to amplify the initial price shock. It generates precisely the kind of vicious positive feedback loop that destabilises markets.
Such hedging feedback loop is an example of market impact that refers to the degree to which large transactions can be carried out in a timely fashion with minimal impact on prices, illustrative of market imperfection, since theories of efficient markets typically assume no market impact as supply matches demand perfectly. As a result, managing risks for large traders requires amending the traditional Black-Scholes-like pricing and hedging model, by introducing an explicit market impact function depending on the number of stocks held by the large trader, since the cost of placing one large order to close a position becomes far greater than the sum of infinitely small orders differed in time.
The optimal execution strategy can then be determined by a parsimonious and realistic no arbitrage agent-based model that endogenously incorporates the market impact of the large traders’ hedging activity (hedging feedback loops), which translates into a fully nonlinear delta hedging strategy. Solving such numerically unstable nonlinear scheme, with significant accuracy and flexibility while keeping stability, needs specific, adequate numerical implementation based on FBSDE techniques. Using such a framework, a large player can then take into account those positive hedging feedback loops in dynamic hedging.
As the major part of derivatives transactions are still OTC, digital technologies offer ways to integrate hedging feedback loops within investment strategies beyond the approach developed above. The use of digital technologies like data virtualisation, robotics process automation, and AI is key to get access and analyse the appropriate proxy data (put-call open interests ratios, gamma ratios distribution, short Futures positioning, equity puts gamma and vega positioning).
