A recent trend across some structured products markets is the increase in number of dispersion products that have been issued. One of these, issued by Credit Suisse, was named as SRP APAC Deal of the year for 2020. Other investment banks active in this area include HSBC and Goldman Sachs. As with many product types, they can be issued as warrants for direct exposure at a reduced premium or turned into a note. The source for all the product details for this article is www.structuredretailproducts.com.
There are different ways to put dispersion warrants together with the name encompassing a broad range of definitions with similar common properties. The simplest product of the three examples is the Credit Suisse warrant issued in Hong Kong. This features 10 stocks (Tencent, JD.com, Nvidia, Amazon, Apple, Goldman Sachs, Johnson and Johnson, Intel, Boeing and Coca Cola) and has a three year maturity. At the end of the product term, the basket average performance is calculated, this is then compared against each individual stock performance, calculating the absolute difference in each case and taking an average. This defines the dispersion measure. It will be at least zero and can only be zero if all stocks show identical performance at the three year point. The warrant itself is effectively a call option on the dispersion amount with a strike of 33%.
Dispersion is a measure of spread and calculates the absolute differences of individual components against the basket performance. For this warrant to payout, the dispersion must exceed the 33% strike level, that is, on average the stocks must be more than 33% away from the basket performance. This would represent a significant variation in stock performances against the basket.
Dispersion warrants make sense for investors who believe that a group of stocks will move significantly differently over time. Dispersion can be thought of as an anti-correlation measure, if stocks exhibit low correlation, we generally expect this dispersion amount to be higher. In certain market scenarios investors may take the view that increased dispersion is likely. This will be driven by a combination of higher volatility and lower correlation because dispersion captures covariance effects. Therefore, the ideal market conditions would be realising higher volatility and different stock, sector or country performances. This almost exactly describes what has happened in 2020.
Another feature of dispersion products is that they are broadly market neutral because the dispersion calculation eliminates overall performance (positive or negative) of the basket or index.
The second example is the HSBC warrant which features the Eurostoxx-50 and S&P-500 indices and the EEM ETF which tracks the MSCI Emerging Markets index. This warrant has a 21 month maturity and its payoff is derived from a measure of dispersion which uses variance swap calculations. The variance swap is calculated for each of the three underlyings and averaged (with a 40% weighting to the Eurostoxx-50). This is then compared to the variance swap for the basket of the three underlyings in the same proportion.
The idea is similar to the Credit Suisse version but a little more complicated in construction as it does not just rely on measuring final performances. If the correlation is low between the three indices then the variance result of the basket will be significantly lower than the average of the components resulting in a higher payoff. The warrant itself pays a geared capped and floored multiple of the dispersion calculation.
The definition of dispersion can be thought of as anti-correlation and the HSBC version captures this notion a little better because the Credit Suisse warrant will only pay out if the final value of the stocks against the basket shows significant dispersion at the maturity date. The HSBC warrant measures volatility (actually variance) through the life of the product and will not depend on the final level alone.
This important difference between these two products trying to capture the same effect is similar to the difference between a volatility swap or measure (like the VIX) and buying a straddle (call and put together). The straddle will pay out if the underlying has moved significantly up or down, a property that is associated with high volatility, however it is possible for an underlying to have high volatility over a period but actually finish near its starting point at a single time in the future. In this case the volatility measure would be high but the straddle return near zero.
The final example is a Goldman Sachs note issued in the US representing delta 1 exposure to the Goldman Sachs US Dispersion Series 31 Excess Return Strategy. As stated in the literature, The Strategy aims to notionally replicate exposure to the correlation risk premium by selling in the S&P 500 Index option market and buying in the single-stock option markets for constituent stocks in the S&P 500 Index with a delta hedge overlay. There are also additional trades made as part of this strategy. A proprietary index approach is inevitably less transparent than both the warrants already examined but it does have the advantage of being open ended. This is more flexible for direct investment and it enables a track record to be built up.
As well as investor appetite for dispersion products, banks are keen to issue them as they can use them to hedge existing correlation positions they will expect to have. For multi-asset products on typical trading books, the two most common groups of correlation dependent products are worst-ofs (for example reverse convertibles and autocalls) and basket linked participation notes. Using worst-of for capital at risk enhances yields, and baskets in participation notes reduce volatility and therefore boost participation. Both of these common but fundamentally different types of trades gives banks correlation risks in the same direction – if correlation increases during the life of the product the bank will suffer mark to market losses. Selling dispersion products gives banks exposure in the opposite direction, since if correlation increases, the dispersion products go down in value. This is a valuable hedge for their business pipeline, and since the overall size of dispersion issuance is likely to be relatively small compared to the standard product types, investors may well be offered attractive levels as the bank will prioritise risk reduction.
A version of this article has also appeared on www.structuredretailproducts.com