Learning Curve – Trading dividends as an asset class
In a negative interest rates environment, dividend strategies are becoming increasingly popular as a way to create positive yield for investors.
By Eric Benoist, head of equity derivatives strategy, Natixis
Dividend trading isn’t new. For as long as there’s been an option market, equity traders have been looking to hedge their residual dividend risk in the form of “synthetics” — simple combinations of calls and puts with identical strikes. The strategy worked, but was typically polluted by repo and rate parameters.
This also involved trading whole strips of future dividends until expiry, as opposed to isolating the exact part of the curve that you needed to keep or remove. Alternatively, you had to enter multiple long-dated “jelly rolls” (forwards calendars) to fine tune your exposure, which could quickly become expensive when every dealer in the market accumulated the same type of exposure and liquidity dried up.
At the beginning of this century, pushed by the phenomenal success of their structured product activities, banks had begun looking for new ways to recycle their “epsilon” — the sensitivity of their option books to dividends — and turned to hedge funds.
However, synthetics’ cumbersome meant that there wasn’t a huge appetite for these structures on the client-side. In fact, it was this initial mismatch led to the birth of dividend swaps. Using those instruments, investors could take a pure static directional bet on the actual level of realised dividends, without wasting time re-hedging their position.
Convertible bond funds were structurally short dividends, and were the first to show interest in buying these assets. A few years later, a market for options on realised dividends emerged and complexity began evolving. But these products could only be traded over-the-counter and volumes remained limited, until Eurex offered its first listed EuroStoxx dividend futures in 2008. In 2010, the exchange followed that offer with the first listed single-name dividend future contracts, and the first listed EuroStoxx dividend options.
With added transparency, success became inevitable and European dividends began to attract a more diverse set of investors looking at the product through different lenses. Directional equity players would look at short-term dividends as a discounted “equity proxy”, reacting to similar drivers but with lower volatility than the underlying stocks, Or they would focus on long-term dividends as a cheaper way to hedge against tail risk (long-dated dividend futures have a higher beta to the downside). Fixed income players would look at dividends as a diversifying source of yield with negative correlation to the rest of their portfolio or as a smart hedge against inflation.
A similar observation can be made about the S&P 500. Robert Shiller’s data shows that since 1900, there have been very few three year windows where its dividend per share dropped by more than 15%: the 1929 crisis, World Wars I and II, and the late 2000s subprime crisis.
Even so, in an environment characterised by negative interest rates and nearly non-existent inflation, it is legitimate to wonder if there are still compelling reasons to remunerate shareholders at 3.5% a year on average.
To answer this, it is worth considering Japan. Despite cultural and economic factors that are unfavourable to shareholder remuneration (ie, rigidity of governance style and inflation struggling to recover to the Bank of Japan’s target, respectively), the Topix payout ratio has fluctuated between 30% and 35% since 2013 and the index’s dividend yield has improved from 2% to 2.4% over that period.
Demographic changes partly account for this phenomenon. For a pensioner, a dividend is available income and the underlying share a future source of income. Therefore, the dividend is part of disposable income, whereas prudent wealth management dictates retaining an equity portfolio for as long as possible.
The Abe government has encouraged citizens to engage in pension planning to supplement government pensions, and, since 2014, it has also been pushing for an improvement in governance standards, including requiring boards appoint external and independent directors. All these measures are favourable to dividend growth in the country.
In Europe, pension systems differ from one country to another. But ageing populations are a reality and models based on the welfare state are reaching their limits, so the risk that dividends might taper towards zero appears a slight one.
In addition, a stable dividend policy benefits companies, as it demonstrates confidence in future activities and the solidity of balance sheets.
So, ending dividend payments would simply be absurd in many cases.
Between now and 2022, dividend futures are essentially driven by companies’ fundamentals. The dividends paid next year will, mostly, reflect this year’s activity. But business visibility is generally good enough for analysts to make robust cash flow assumptions 18-24 months into the future and produce dividend estimates matching that horizon.
Analysing each company’s prospects in the EuroStoxx, we estimate that 2021 will bring a total of 125.8 dividend points (very close to the Dec21 future’s mark of 123.9), 17.5% of which will come from a fragile but resilient banking sector. Clearly, remaining upside in the front part of the curve appears limited.
Arguably, more value can be found in the back end, beyond three years, where it becomes more difficult to build a fundamental view and dividends mainly react to long-dated option flows.
Each market is different. In the US, where protection buying flows from pension funds and insurance companies represent a large chunk of the volume, the S&P 500 dividend curve is upward sloping out to Dec30.
But in Europe, where banks’ structured products hedging flows often prevail and downside volatility tends to be heavily sold as a result, the opposite is true. Interestingly, this inverted term structure tends to steepen whenever equity markets experience turbulence.
From a bank’s perspective, autocall products can be analysed as the sum of a long “Down & In Put” and a short digit call exposure. The end client receives a coupon, as long as the spot remains below the recall barrier and experiences capital losses only if the spot breaches the downside barrier. This results in a short delta and long dividend positioning.
When markets go down and autocalls become less likely to be recalled, dividend sensitivity increases with the delta of the puts and selling pressure increases in the longer-dated tenors, as dealers start to offload their dividend risk. The lack of liquidity beyond four years exacerbates that effect and can lead to very dislocated dividend levels, offering compelling entry points and opportunities.
Alternatively, “Pull to Par” strategies (systematic or not) will offer to buy discounted long-dated dividend futures and carry them until they start converging towards their front year’s fundamental value. In extreme circumstances, the mark-to-market of these positions can be volatile but implied dividends normally start converging 18 months before expiry, thereby providing a “soft floor” to investors.
Over the past few years, financial engineers have done a lot of work on dividends and there are many ways to implement views and extract positive yield in a world where generating income has become increasingly difficult.