Structured product sellers and div swaps

From a Barclays primer on dividend swaps:

We note that for shorter periods of time, implied dividends can be more volatile than spot as dividends often trade away from fundamental value for technical reasons (as the structured products sellers become longer implied dividend risk as spot declines, and they hedge this risk by selling dividends, which can cause implied dividends to over shoot on the downside).

Why do structured product sellers become longer implied dividend risk as spot declines? I have clue how this mechanism works.

• Hi Permian, could you possibly add the link to the paper? Seems rather odd, even if you consider selling a European call under BS this doesn't seem to be the case (it's Friday though so maybe I made some mistake). – Quantuple Feb 23 '18 at 14:53
• trading-volatility.com/… – Permian Feb 23 '18 at 15:35
• Ok no real added info in the document. In my opinion such a statement cannot be made in a general. For instance, when selling a standard call you need to buy back the dividends you sold to hedge the short forward exposure at inception when the spot price declines (since the exposure is then less than before, consider the case where the spot declines so much that the option becomes completely OTM). You can also compute $\partial^2 C/\partial S \partial q$ for a call under BS to convince yourself. – Quantuple Feb 23 '18 at 17:32

The paper is generally correct, but it is not a general statement, as in a general truth of options hedging in a theoretical context, rather a statement regarding how the structured derivs market is typically set up: retail and institutional investors buy a large number of products that at their core entail the dealer buying (from the investor) long-dated (3y+) otm put options, or perhaps even more often, down-and-in put options. The idea is that the value of this potential downside loss for the investor can be paid back by the dealer in the form of better returns (eg larger coupons) than what a simple principal-protected product would achieve. A typical popular product is a so-called Autocallable Note (see elsewhere).

Schematically, these products make dealers longer (implied future) dividends (ie. shorter delta to the forward) as the market goes down and one way to hedge that risk is to sell long-dated dividends via div swaps (or div futures).

The effect is sometimes quite violent as was seen on Eurostoxx div futures in 2008-9. The divs moves will then largely overshoot that of the index itself. In such instances it is fair to say that the forward div curve does not reflect “the expectations of market participants”, rather it reflects the large-scale forced selling by a number of them.

As a side note and this is the same effect at play, you also see that the typical term structure of div futures is downward sloping which opens up interesting opportunities if one can weather the swings in the meantime.

• Its not clear to me how this works ""these products make dealers longer (implied future) dividends (ie. shorter delta to the forward) as the market goes down – Permian Feb 25 '18 at 10:49
• Not sure what this means, "The divs moves will then largely overshoot that of the index itself." – Permian Feb 25 '18 at 10:49
• "typical term structure of div futures is downward sloping which opens up interesting opportunities", why should dividends be upward sloping? – Permian Feb 25 '18 at 10:50
• If you’re long a put, then your exposure to future dividends increases as the spot price goes down, since your (short) exposure to the underlying forward itself increases. To your second point: moves in implied future dividends may be larger than moves in the index, thereby implying a decrease in yield. Wrt to term structure, you would think companies tend to try to increase divs over time, although I agree it’s not an iron law of economics. Certainly there is no reason either that it should be downward sloping but it tends to be. – Ivan Feb 25 '18 at 13:25
• @Ivan thanks for your answer it now makes sense to me as well what they meant. – Quantuple Feb 26 '18 at 11:45

To add to the above on a more practical note:

In general, SP desks make money on the individual product when the underlying declines. Dividends make the underlying decline, hence they are naturally long dividends.

Take an auto-callable product which is exercised if the spot is above a pre-determined strike each year and say the SP desk sells this structure over 5 years. Once a year, on a pre-determined date, the product will be exercised if the final price is above the strike. Maximum length of the product = 5 years, minimum = 1 year.

If the market drops, the probability the product fails to be exercised increases, and hence the expected term of the product increases. If they were long dividends before, now that the market has dropped and the expected lifetime of the product has increased, their long exposure to dividends will increase.

Just to add a remark on top of Ivan’s excellent answer, note that the core reason IB package those KI puts in the autocallables and other SP for investors is not just in order to make the coupons more attractive to the investors but fundamentally to buy back the volatility skew that the vanilla desk is structurally seller of.