A Normalising Volatility Regime?

Volatility and Tail Risk Educational Event – London 2018

Hamlin Lovell
Originally published in the May 2018 issue

The annual Volatility and Tail Risk Educational Events (held in Sydney, London and Zurich) have raised more than $500,000 for charities such as VIMBA, Woman’s Trust, Vision for a Nation and Beanstalk. The London event held on April 12th, 2018, generated $106,000. The Hedge Fund Journal nearly always attends the London day and here we review highlights from a very candid discussion amongst leading hedge fund managers of volatility and other strategies, economists, exchanges, brokers, allocators, academics and other market participants.

Volatility flashpoints

There are always Cassandras calling for crises, and a stopped clock must be right twice a day, but in early 2018 it seems that a range of risk factors are coalescing to ramp up volatility.

Event chair, Maya A.Z. Rodriguez, Managing Partner of AZR Capital, spotlighted that geopolitics and domestic politics are heating up, with far-right politicians winning power in multiple countries while the Middle East faces dictatorships, civil wars and chaos. Veteran war reporter, Robert Fisk, said he is investigating the provenance of weapons used by Isis and Nusra. His findings could heighten political tensions outside the region.

Non-financial debt as a percentage of GDP has surpassed pre-crisis levels in China and the G20.

Maya A.Z. Rodriguez Managing Partner, AZR Capital

Trade wars are another risk, pointed out by Lena Komileva, Chief Economist at G+ Economics, who also reckons that “markets are complacent about central banks ending quantitative easing for the first time in a decade. The Fed is withdrawing $1.5 trillion of QE, of which $380 billion is coming this year. The ECB’s QE policies have had effect through multiple channels – credit, currency, interest rates, duration, and signalling” – all of which could back-pedal. “February 2018 saw financial conditions turn negative for the first time since February 2015 which preceded a major stock market correction, and the IMF’s Financial Stability Review has identified medium term vulnerabilities rising,” she added. Higher rates could have greater impact because “non-financial debt as a percentage of GDP has surpassed pre-crisis levels in China and the G20,” she mapped out.

Faster inflation also seems probable. “The output gap has already closed in the US and Germany and may do soon elsewhere in Europe. A whole host of factors that have kept inflation low – labour reform, globalisation, little unilateral currency devaluation, technological innovation, demographics, and a cyclical undershoot in the jobless recovery no longer apply. Ten-year break-evens on inflation linked debt could easily move up by a percent,” said Komileva. She judged that “the focus of monetary and fiscal policy is no longer volatility control, and the regime has changed from euphoria to anxiety”.

Artemis Capital Management LP founder, Christopher Cole, contends that the sheer volume of institutional money positioned for short volatility has begat lower volatility in a reflexive cycle that could easily upend into an upward spiral for volatility. He believes that “$2 trillion of financial strategies are explicitly or implicitly short of volatility. Only $60 billion is in explicit short volatility strategies, but other strategies including risk parity, variance control and risk premia also replicate the dynamics of short volatility”.

Cole also views equity investors as complacent, and claims that “US equities would be in an earnings recession without share buybacks – which also reduce volatility”. Societe Generale sees risks of a recession by 2019 or 2020, and thinks that the profit cycle could be peaking (this view was not shared by all speakers; Caxton Associates’ Associate Partner and Portfolio Manager, Vishnu Kurella, sees “no indication of recession around the corner”).

For Cole, “the market action seen in February 2018 – triggered by a mere 40 basis point rise in rates – is only an appetiser for a main course of further equity volatility should rates spike 1% to 2% higher as they have in other reflationary periods”. Cole expects that an inflation shock could be the trigger factor.

How to mitigate tail risk?

Volatility strategies per se are among many that may profit from a more volatile climate or even provide tail risk insurance. US pension fund CalSTRs allocates to a range of ‘risk mitigating strategies’ that are expected to improve returns, reduce volatility and reduce drawdowns.

Aberdeen Standard Investments looks at tail risk from many angles, with the objectives of “avoiding catastrophic losses, mitigating negative skew, and maintaining liquidity to take advantage of opportunities in stressed markets,” said Senior Investment Manager, Stephen Coltman. Four currently popular tail risk strategies that Coltman critiqued included trend-following, covered calls, collars, and fixed income. He pointed out that “trend-following gained in popularity after strong performance in 2008 but can struggle in a fast moving crash scenario, covered calls sacrifice the right tail to partially cushion losses in a sell-off, zero cost collars require active management and pricing of skew means you cut off the right tail more than the left tail, and long duration is vulnerable to a change in correlation between bonds and equities.” Coltman is of the opinion that “a portfolio of strategies including risk premia with low or negative correlations to equity markets, trend following strategies, and long volatility strategies, can be combined to deliver an attractive risk diversifying return stream, particularly as the costs of accessing such strategies are coming down”.

Coltman flagged up that fixed income may not work as a tail risk hedge in the future, if negative correlations between equities and bonds break down, and this sentiment was echoed. David Dredge, CIO of Convex Strategies at City Financial Investment Company, agreed that “fixed income is no longer an efficient way to mitigate risk, and that correlation breakdowns show the weaknesses of mean/variance optimisation”. Eric Bissonnier, CIO of LumX, concurred that mean/variance optimisation is not really useful and that investors need something to help with real risks in the portfolio, while mitigating the cost of carry. Christopher Cole observed that the overriding problem with bonds in 2018 is that they have so little headroom to rally. To match historical crisis performance “fixed income yields would need to plunge as low as -3% or -5%, which is highly improbable. It is much more likely that bonds and equities decline together,” he said.

Volatility strategies: a diverse spectrum

Homing in on volatility strategies, their exposure to directional volatility can be long, short, neutral or variable. Only a subset of volatility strategies (a minority after a nine-year bear market in volatility) are long of volatility, and only a further sub-division of those are seeking to provide meaningful tail risk insurance. Membership of a tail risk index (such as the CBOE Eurekahedge Tail Risk Hedge Fund Index) merely means that a strategy satisfies that index’s criteria for inclusion, which can be as simple as pursuing a long-only volatility strategy. An added nuance is that even where strategies are long-only volatility, whether and to what extent they may profit from a spike in volatility depends on rules or discretion (and market liquidity) that determine how and when gains are monetized. This may explain why the aforementioned index reported negative performance in February 2018.

One example of a long-only volatility manager is 36 South Capital Advisors, whose CIO and CEO, Richard (Jerry) Haworth, says, “we make no apologies for negative carry”. Haworth reminded attendees at the event that tail risk protection and negative carry are two sides of the same coin. Until a tail event occurs, negative carry is a necessary by-product of providing true tail risk protection in order to achieve convexity in a crisis.

At the same time, Haworth noted that a portfolio can be designed to be long volatility, whilst also mitigating carry costs, but this necessarily reduces the expected return in a crisis.

While 36 South is known for being a tail risk manager, Argentière has an absolute return objective. As well as being long volatility, Argentière pursues relative value and volatility arbitrage strategies, such as index dispersion, intended to generate positive returns in normal market conditions.

Non-directional dimensions of volatility

Indeed, the absolute level of volatility is not a return driver for all volatility strategies. February 2018 saw a big directional spike in volatility, but other dimensions of the volatility surface, such as skew, term structure or regional differences, are more relevant for some arbitrage trades. And although outright volatility remained above recent averages in April 2018, “volatility of volatility has come in a lot, without the same rebalancing flow, and convexity is now quite cheap in the US,” noted Vishnu Kurella.

A regime change in volatility?

There was some debate over whether the jump in equity volatility was likely to mark a regime change. Maubourguet argued that “the short volatility trade is a self-fulfilling feedback loop, which has been damaged. The gamma positioning of ETFs has now gone. The short volatility trade is not going to return with the same size or confidence any time soon”. Pierre de Saab, Partner at Dominicé & Co, reflected that “the demise of the short volatility trade had to come at some point, as it was a classic behavioural bias and banks like Societe Generale had pointed out the crowding effect”.

The February 2018 shock was isolated to equity markets, with no contagion into other asset classes. Indeed, “the VIX is a tiny percentage of the global volatility market, and long-term interest rate volatility actually went down in February,” observed Haworth. Kurella agreed that “rate volatility and FX volatility are close to the lows”. Haworth added that tail risk strategies aim to capitalise on a systemic surge in volatility across all asset classes, and are not designed to make supernormal returns in smaller events, such as February 2018. Haworth summed up that “every volatility regime starts with a volatility spike as we saw in February, but not every spike leads to a higher volatility regime”. David Dredge suggested that each cycle differs, but was confident of one thing: “there will not be another round of central bank QE and monetary accommodation this time”.

Liquidity bottlenecks

Unfortunately, the most marked regime shift identified by some speakers in February 2018 was impairment of liquidity. This is a practical issue to grapple with when implementing many volatility strategies. It may be more costly, difficult or even impossible to exit and monetise some positions (at least immediately) after a volatility spike. “Lack of liquidity is the new leverage. Due to share buybacks trading volumes and free float shares in the S&P 500 index are now at levels last seen in the late 1990s, and this is the only bull market in history with lower and lower trading volumes. What people don’t realize is that February 2018 was an illiquidity event rather than a volatility event triggered by rates. In February, exchange traded products traded 2-3% below their liquidation value. There was no price on variance swaps. Option spreads went from five cents to three dollars wide,” claims Christopher Cole. Vishnu Kurella also noticed “spreads blew out from 50 cents to 5 USD wide”. Will Bartlett, CEO of Parallax Volatility Advisers, agreed: “liquidity was as bad as we have seen for 23 years, as much buying over a short timeframe created a squeeze. It was a challenge to find liquidity during the episode”.

Market structure

Liquidity is partly a function of market structure. Societe Generale’s Managing Director and Global Head of Flow Strategy and Solutions, Financial Engineering SG CIB, Kokou Agbo-Bloua, posed the question of whether volatility traders believe that algorithms and machines, or other human traders, are on the other side of their trades. “In the last few minutes of VIX trading on the screen, it feels like you are trading against a machine,” said Pierre de Saab. Bartlett explained that “the US market is very fragmented amongst 12 exchanges, which internalise the retail order flow, making it very tough for us to interact with retail order flow”. In contrast, Europe is perceived as a more institutional market. Eurex’s Global Head of Sales for Equity & Equity Index Derivatives, Markus Alexander-Flesch, claims that “the VSTOXX contract is mainly used by professional traders”. He also argues that “the European market is structured very differently from the US market as regulations encourage transparency”. In Asia, the scale of structured product issuance is such that it proliferates across institutional and private capital, according to Dredge, who sits in Singapore.

Evolving analytical tools

Hedge fund managers, including volatility managers, are using advanced techniques not only for algorithmic execution but also to implement strategies. Techniques include game theory, AI, Big Data, and Natural Language Processing (NLP). Some of these techniques may be over-hyped, but all of them can provide useful inputs and insights for various parts of the investment process.

W. Ben Hunt, Chief Investment Strategist of Salient Partners, and author of Epsilon Theory, explained “game theory does not predict and won’t make regressions run better, as almost every game has multiple equilibria, outcomes and balancing points – none of which can be predicted. But you can track the dynamics of the game”. He also finds AI useful “to help you observe, predict and visualise the dynamics, for unstructured data”.

Javier Rodriguez-Alarcon, Head of QIS EMEA Asia ex-Japan, GSAM, agrees that AI techniques such as Natural Language Processing or NLP can help to address the challenges of analysing vast amounts of unstructured data like text: “the volume of unstructured data, such as pictures, text or videos are growing much faster and significantly overtaking structured data, which is numeric and can be easily analysed by computers”.

Cole finds “machine learning is an incredible tool to find solutions for closed systems, such as the game of AlphaGo, or self-driving cars, where the rules do not change as the game progresses – but markets are not a closed solution”. For Cole, “the drivers of volatility are very consistent over the past 100 years, the drivers of 1928 were very similar to 2008. Machine learning won’t tell you the spark that leads to the fire, but does give insights into datasets that you would not otherwise get, so is very valuable for brainstorming”. UK pension fund quant portfolio manager, Tony Guida, is more enamoured of AI and has found “econometric models were all about reducing in sample bias but now machine learning is revolutionary for prediction out of sample, and this is what finance should be doing”.

Hunt summed up “AI lets us ask better questions about markets and patterns. It does not replace humans – it saves them”.

Trade ideas across asset classes

“QE paradoxically turned out to be deflationary, and tightening could prove to be inflationary,” pointed out Rich Haworth of 36 South, who views commodities as a Cinderella asset class, with most retail and institutional investors having no allocation. 36 South sees potential for “a breakout in hard assets” and view gold options as cheap versus those on Amazon stock.

For de Saab, “the initial volatility shock lasted longer than expected, with volatility staying in the 20s when we expected it might calm down to the mid-teens. We are finding good risk reward, without predicting the level of interest rates and volatility. Convexity on the S&P 500 is cheap”. Kurella agreed that S&P 500 convexity looked interesting.

Will Bartlett remarked that “the regime has changed with realised volatility above implied volatility, but many investors are continuing with the same strategy”. Parallax has instead “shifted from the S&P 500 to a short Nasdaq volatility stance, to profit from the spike in implied volatility as investors buy huge amounts of put options”. Bartlett also noted that “Euro stoxx forward volatility got very depressed as did gold volatility”. Bartlett has as well noticed volatility shifts now occurring in days rather than weeks, so he needs to rotate exposures more frequently to keep pace with shifting relative values.

Kokou Agbo-Bloua jested that some funds might not be around next year, if a volatility blowout causes them to meet the fate of ETF XIV.