Like postmodern literature that eschews conventional plot, narrative, and characters, financial markets have become increasingly abstract and remote. Volatility trading offers a good example of why this is so dangerous.
In February, Goldman Sachs Group Inc. reportedly made $200 million in profit on a single day from volatility trading. Anticipating increases, Goldman purchased volatility then sold it profitably to investors needing to cover short positions when the Cboe Volatility Index surged by more than 100 percent and the S&P 500 Index fell by about 4 percent. Those earnings, which were matched by losses elsewhere in the market, were large when one considers that profit at the bank’s entire trading desk exceeded $100 million on only four days in the previous year.
Volatility loosely measures asset-price fluctuations, expressed as the standard deviation of price changes. The problem is that its underlying statistical assumptions rarely reflect reality. Volatility can be measured with certainty only ex post. Even then, different methodologies yield different results. Trading involves speculating on future volatility, measured by implied volatility, which is itself extracted from traded option prices. In a circular process, traders input expected volatility to calculate the option price from which implied volatility is determined.
Implied volatility is in practice, unsurprisingly, a poor predictor. It’s impossible to estimate accurately the impact on asset prices of political developments, natural disasters, wars, pandemics, or financial events such as the restructuring of the euro, a significant slowdown in China, trade wars, or unexpected growth or inflation figures.
Yet this fundamental inaccuracy hasn’t prevented volatility from being converted into something tradable. VIX futures, over-the-counter variance products, and option-selling may total as much as $100 billion. Volatility models also drive trading strategies totaling as much as $2 trillion. Risk-parity or variance-control funds target a set risk level defined by volatility. Long-short strategies, as well as algorithmic and momentum trading, require volatility inputs. Firms engaged in stock buybacks indirectly write put options on their own shares. Central banks implicitly sell put options in underwriting market values and liquidity.
Volatility is also a key input in financial risk models that determine compliance with risk limits, capital requirements and collateral arrangements. Since 2009, the capital requirements of regulated financial institutions have increased. But central banks have simultaneously suppressed volatility, understating risk and risk-weighted assets, paradoxically lowering the amount of capital required.
Assuming that central banks will maintain stability, investors trade off small gains against the risk of substantial losses. In the typical cycle, betting against increases forces volatility ever lower, inducing mispricing whereby risk-takers are inadequately compensated. Problems occur when volatility increases or becomes unstable due to a shock.
Any breakdown is rapidly transmitted. Investors who have sold volatility for premium income and are unhedged need to sell underlying assets to rebalance their portfolios. Amounts traded increase non-linearly, with progressively larger amounts having to be sold if prices fall sharply. Price fluctuations in one market spread across asset classes as all volatilities are re-rated. As most trading is done on a leveraged basis, margin calls accelerate the rebalancing and contagion as everybody seeks to increase cash levels. These problems are compounded by rapid declines in trading liquidity. Opaque prices, combined with conservative mark-to-market levels by counterparties seeking to ensure adequate security for positions, will exacerbate decreases in values.
In the financial economy, significant losses can result: In February, a dislocation in a roughly $50 billion sector of the market helped trigger global losses of some $5 trillion.
Higher volatility also increases measured market risk. Equity volatility is used in credit models to calculate the probability of default, meaning that higher levels increase credit risk. Capital to be held against assets and the amount of collateral required for trading rises, increasing the cost of trading and limiting the ability of banks to make markets and warehouse risk. Funding becomes expensive and the supply of loans to businesses and leveraged investors falls, worsening the problem.
In the real economy, increased volatility heightens uncertainty, forces up the cost of capital and reduces available funding, resulting in lower consumption and investment. Defaults and credit losses then feed back into the financial economy via the banks in a familiar pattern.
In most industries, complex or dangerous equipment must be extensively tested and deemed fit for purpose before it can be used. In contrast, policy makers tend to view financial products as providing a self-evident benefit. This encourages investors to trade obscure variables in the hope of profit. The utility of these instruments is rarely questioned even when, as in the case of volatility trading, they distort the financial system and increase the likelihood of future crises.
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
To contact the editor responsible for this story:
Timothy Lavin at email@example.com