Avoiding dividend meltdown

Dealers are starting to pay closer attention to dividend risk housed on their exotic books after many incurred sizeable losses last year. What are banks doing differently and can another dividend meltdown be avoided? Matt Cameron reports

Until last year, dividend risk was not thought to pose any major risk to the exotic books of structured products issuers. Quite the opposite, in fact - it was seen as a boon. Most dealers were long dividends as a result of structured products sold to retail and private banking clients. With dividend payouts expected to increase steadily over time, this exposure was seen as an easy way to bolster profits. "Dividends were a facile way for dealers to make money," says one London-based equity derivatives strategist.

Perceptions were violently shattered in the first quarter of 2008, when a spike in correlation and volatility combined with collapsing dividend expectations to deal a heavy blow to exotic equity derivatives books (Risk July 2008, pages 21-241). Dealers were hit again after the collapse of Lehman Brothers last September, which triggered a freefall in equity markets and a further sharp contraction in dividend expectations (Risk December 2008, pages 21-232).

With dealers facing hefty mark-to-market losses, many rushed into the dividend swap market to hedge their exposure. Investors, too, were hurt on long dividend trades, while hedge funds in particular were forced to unwind positions amid a mass deleveraging.

With most firms all positioned the same way, implied dividends on European stocks, as derived from the prices of dividend swaps on the Dow Jones Eurostoxx 50 index, were crushed amid the weight of selling pressure. On March 10, 2010 Dow Jones Eurostoxx 50 implied dividends were at 54 points, having traded at 134.4 points when Lehman collapsed on September 15, according to figures from Barclays Capital. Similarly, 2011 implied dividends crashed from 134.5 points on September 15, 2008 to hit 51.7 points on March 10. Over the same period, the Dow Jones Eurostoxx 50 index fell from 3,151.17 to 1,919.53.

However, implied dividends have recovered in recent months, reflecting a rally in equity markets and realisation dividends may have been oversold. The Dow Jones Eurostoxx 50 index rose to 2,509.22 by June 12. On the same day, 2010 Eurostoxx implied dividends had climbed to 79.9, while 2011 implied dividends on the Eurostoxx had increased to 74.4.

Dealers say they managed to square a large chunk of their dividend exposure earlier in the year, easing the forced selling pressure. In addition, the sale of new structured products have slowed markedly in 2009, creating less long dividend exposure for banks.

Demand for products such as reverse convertibles and auto-callables, in particular, have been hit with the decline in the equity markets. These pay investors an enhanced coupon so long as the reference stock price does not fall below a predetermined barrier. If the stock does fall below this level, the product redeems in shares. These structures mean the dealer is long a knock-in put option and long dividends. As equity markets tank and approach the barrier, the dealer would need to hold more of the stock as a delta hedge, further increasing the long dividend position. With less of these products being sold, and with many of the legacy products having already knocked in with the decline of the equity markets, long dividend exposure is now well below the risk parameters of most banks.

Dealers note that implied dividends are returning to more fundamental levels after the mass sell-off earlier in the year. However, the market still has a pessimistic view on future dividend growth: as of June 12, 2010 implied dividends were trading at a 30% discount to 2009 implied dividends.

Pressure on bank exotic books may have eased, but the losses experienced by many remain clear in dealers' minds. Are banks looking to alter the way they manage dividend risk as a result?

Fundamentally, little appears to have changed. Dealers say the focus has been on improving liquidity of index dividend swaps, putting greater focus on the liquidity of underlying stocks when taking on single-stock dividend exposure, a reassessment of models used to forecast future dividends, and more conservative dividend assumptions when pricing structured products.

"Markets falling and companies cutting their dividends is not extraordinary and has happened in the past. What has been spectacular in the past year is the extent of the move," says Eric de Pommerol, head of delta-one trading at JP Morgan in London. "However, things have not changed dramatically on the risk management side and how the hedging of dividends is dealt with. Risk management has always been a function of liquidity, and ultimately dealers will tailor their risk to the liquidity available in the market."

The risks posed by dividend exposures on single stocks, in particular, have attracted attention. These have been built up due to popularity in recent years of structured products referenced to baskets of stocks. Given low levels of liquidity in the single-stock dividend swap market, dealers tended to hedge on a forward basis by buying a put option and simultaneously selling a call on the underlying stock. However, a sharp decline in liquidity in the options market at the height of the crisis made it difficult and expensive to hedge these exposures. Alternatively, index dividend swaps could be used as a proxy hedge - but this created significant basis risk for banks.

"The problem in most cases is the basket has been created with stocks that tend to have dividend yields higher than the index itself. So there were often considerable discrepancies using the hedge," explains Arie Boleslawski, head of exotic equity derivatives trading at Societe Generale Corporate and Investment Banking (SG CIB) in Paris. "What is more, if a stock heavily underperformed, it could be removed from the index, which would affect the hedge negatively. Being long stocks and short the index became worse with the dispersion of the basket."

As a result, dealers say they are more wary of proxy hedges. Many are also reluctant to price single-stock and basket trades that include illiquid shares. Demand for these types of products has waned considerably since the sell-off in equity markets, with investors tending to focus on index products. For those single-stock trades still being conducted, banks are being a lot more conservative in their pricing and dividend analysis, with some capping their assumptions on dividend yields.

"The major change in our approach towards dividends is the degree of aggressiveness towards the pricing of single-stock dividends," says Moritz Seibert, head of exotic equity pricing at Royal Bank of Scotland (RBS). "We are taking a much more conservative approach when it comes to determining the growth rate of some of these equities, and are capping the dividends of some stocks with different degrees of aggressiveness, depending on the stock and sector. In some cases, financial stocks have plummeted by as much as 80-90%, so their dividend yields increased to levels as high as 10-15%. Therefore, compared with 18 months ago, we are now capping some of these yields at a more conservative yet reasonable level."

Banks are also looking to create structures that reduce dividend exposure and ease risk management. For instance, some firms are marketing products referenced to total return underlyings, which include dividends and assume they are reinvested in the stocks, rather than price return indexes, which ignore dividends. The concept of total return has not been popular in the past, as investors would have to sacrifice some potential upside. The fact dividends are reinvested in the stocks means there is less participation in the upside, compared with a product referenced to a price return basket or index where the dividend flow can be used to buy more options.

However, some dealers are looking to develop products based on total return underlyings, yet which provide the investor with additional upside potential. SG CIB, for instance, has used the concept of a synthetic dividend, where a total return index is adjusted by an assumed dividend yield. This synthetic dividend is then used to increase participation levels. If dividends realise at above the synthetic dividend level, they are reinvested in the stocks. At the same time, the dealer is able to reduce some of its dividend exposure.

"By creating these trades, the client is being given the upside. It also helps to reduce the dividend risk on our side," says SG CIB's Boleslawski. "There is still some educational work to do on these structures, but some clients have shown interest."

Thomas Salter, head of product development in equity derivatives at JP Morgan, agrees: "A synthetic dividend or an excess return approach is used to achieve the client's target exposure on the upside while avoiding dividend risk for the provider."

Banks have also been marketing other structured products that give investors access to dividends. However, rather than a risk management tool to help offset residual dividend exposure, dealers say the primary motive is an attempt to generate interest in dividends as asset class and enable investors to take advantage of current distressed prices.

BNP Paribas, for instance, launched its Harewood Euro Long-Dividends fund in May, designed to trade dividend swaps on the Dow Jones Eurostoxx 50 index. "Initially, the idea behind the fund was to generate broader interest on dividends on the buy side, so customers could benefit from a traditionally imbalanced market with attractive, almost distressed entry levels," says Bertrand Delarue, global head of product engineering at BNP Paribas in Paris. "Like many dealers, our exposure has decreased over the past six months, while the flow of structured products has also slowed since the onset of the crisis. But the hope is the fund is successful and a new investor base of natural buyers develops who are willing to participate in the dividends and hence make the market for dividends more liquid and varied."

JP Morgan has also launched a product that gives access to dividends. The three-year note is linked to the Dow Jones Eurostoxx 50 dividends realised in 2011 expressed in index points. The trade pays a fixed indicative coupon of 5%, with upside exposure to the index 2011 dividend beyond 50% of the 2008 realised dividend. This means the investor receives 100% plus any gains on a dividend call struck at 50% of 2008 dividends minus any losses on a dividend put spread struck at 40% and 32% of 2008 dividends. Investors suffer capital loss for realised dividends below 40% of the 2008 level, with full capital loss below 32% of the 2008 level.

But the dividend products sold have been few and far between and have been structured on major headline indexes, so do not solve the problem of how to hedge single-stock dividend exposure. "Many of the products being marketed are structured on major indexes, while index dividend risk hedges are accessible in the synthetics and dividend swap market," says Gilles Dahan, head of derivatives trading for Europe, the Middle East and Africa at Citi in London. "So the primary reason for these products being sold is not to offset residual exposure but, like any product, aiming to sell a story to the end investor."

Aside from designing new products, dealers have also been attempting to lessen dividend risk by revisiting model assumptions. Typically, dealers use both fundamental corporate analysis of stocks and yield-based models to determine future dividends and hence price options with the derived dividend assumptions.

Yield-based models typically calculate the dependency of dividend payments to movements in stocks. In other words, they attempt to determine how dividends would change in proportion to shifts in stock prices. But there are dangers in relying too much on the outputs of these models. "If you model the dividends proportionally, and the model that determines the dividend is fully proportional to the stock, if you try to recalibrate the forwards, you may find a very big jump between the market price of some instruments such as deep out-of-the-money puts and variance swaps and the theoretical price you get from the model," says SG CIB's Boleslawski. "Essentially, the model is not wrong but there is just no pricing consensus in the market."

However, manipulating and improving existing models does not shield dealers completely from dividend risk. "There have been modelling improvements, but there is no magic formula that allows dealers to be immune from liquidity in dividends," says Citi's Dahan. "You could argue the violence of the moves in 2008 would have been more extreme than what some models would have suggested. Models wouldn't have suggested single-stock dividends would plummet to zero - and this is what happened to implieds on some stocks for a short period last year."

So, can another dividend meltdown and further losses be prevented if the market were to move in the same way it did last year? "The simple capital-protected note - the stalwart of the structured products business - is definitely not going away, and it is hard to see the structured products business changing radically," says one London-based equity derivatives trader. "Will dealers accumulate large long dividend exposures and run the risk of being hit again? It is possible. The past two years have shown there was complacency in the pricing of equity derivative risks, but one hopes those that got burned last year will have learned their lesson."

One sure-fire way of ensuring dealers are able to risk manage dividend exposures is to create an active two-way market in dividend swaps. If this does not materialise, dealers may have to think twice about the structured products they create, concedes Delarue of BNP Paribas.

"The accumulated dividend positions were the result of an imbalanced market. But if we are able to successfully develop a more balanced market in terms of buyers and sellers of dividends, this risk can be mitigated. If the market does not develop, we may have to start altering the structured products we sell."