Auto-callables have been actively traded in Europe for several years and have now become an increasingly influential force in flow derivatives. Initially, they were issued on baskets of stocks, then on indices and now increasingly on single stocks.
Recent activity in Asia, for example in Korea in January last year, suggests structured products play a pivotal role in volatility markets, although the theoretical effect in Europe and the U.S. is diluted because of greater activity from hedge funds, corporates, institutional investors and overwriting funds.
An auto-callable product is a structured product that has an automatic call feature on pre-prescribed dates known as the auto-call dates. These are usually from monthly to bi-annual intervals.
While few auto-callables are exactly the same, most share common characteristics:
1. A large coupon is payable if an automatic call level is triggered when the underlying is above the strike on an auto-call date. If not auto-called, the coupon escalates, often to the next multiple for the next auto-call date; and
2. A sold knock-in put is often embedded to reduce the structure cost. This is referred to as soft protection because the product is usually capital protected unless this knock-in put is triggered.
For illustrative purposes in this Learning Curve we assume the following:
• Three-year maturity with annual auto-call dates and 100% strike;
• 10% coupon escalating to 20% and 30% in years two and three if not auto-called; and
• 70% downside knock-in put with continuous barrier which can knock-in any time.
At each auto-call date there are two possibilities that determine if the product will be auto-called or will not pay a coupon but continue to the next period:
• If the underlying is above the strike (100%), accumulated coupons are paid at 10% per year of life and the whole structure is terminated; and
• If the underlying is below the strike (100%) the structure remains alive and the potential coupon escalates by a further 10% for the next auto-call date.
If at any time the underlying is below the knock-in put strike (70%), the auto-callable holder becomes short the put and this soft capital protection is removed.
Technically speaking, in an auto-callable product the client is short an out-of-the-money knock-in put and long a strip of contingent at-the-money digital options, which pay coupons and knock out the whole product if it is in the money at the auto-call dates.
The most common explanation for auto-callable popularity is the potential for a high coupon, which is attractive in the current low yield environment for bonds, property and equities.
Investors who buy auto-callable products are looking for a long-term equity-linked investment providing a conditional 100% principal protection at maturity and a potential for a large return on investment from coupons. The investor has a view the underlying will have a moderate increase over the life of the product and will not trade below the knock-in level.
Another source of popularity arises from the fact auto-callable products may have a shorter lifespan than other structured products. It is hoped in these situations, investors flush with new cash from the coupon paid, might invest in further auto-call products. Many other structured products lock in an investor's money for a long duration--five to 10 years is common--so for issuers, the prospect of a seemingly long-term product expiring early may provide the chance to issue another product to the same investor.
From an investor's perspective the risks are:
* Early maturity;
* Performance of the underlying--with full capital exposure to falls in the underlying below the sold knock-in strike;
* A rise in interest rates during the investment period may result in mark-to-market losses;
* Liquidity risk and spread costs if an investor seeks to sell their notes prior to the maturity date; and
* Event risk due to some adjustments to the terms of the contract such as mergers and trading suspension.
From an issuer's perspective the risks are:
* A discontinuous valuation pattern near auto-call dates when delta hedging may require a theoretically infinite multiple of the notional to be traded rapidly;
* An indeterminate time to maturity. Vega and gamma exposure is highly dynamic. As the underlying rallies, vega exposure gravitates toward the next auto-call date; and
* Correlation exposure between bonds and the underlying. The sensitivity of the fair value of an auto-call product to this hybrid correlation is negative. This stems from the risk to the issuer that as the underlying rallies, the value of the auto-callable increases but if bonds are also rallying, the discount rate falls and will also serve to increase the product's fair value. This is because the present value of the coupon payment will be higher. In this situation, high correlation has a double negative effect for the issuer.
Auto-Call Effect On Flow Derivative Markets
The delta changes can be significant and irregular for auto-call products particularly when the underlying is trading near a knock-in or knock-out level close to expiry.
The auto-callable product experiences a fair value jump when a knock-in or knock-out occurs. In our example, at the 70% level, the issuer requires a large hedge position so that profit/loss on minor underlying movements are sufficient to offset large fair value changes in the auto-callable.
The chart shows the massive change in delta position required when the underlying of an auto-callable trades close to a knock-in or knock-out level near an auto-call date or expiry.
The auto-callable issuer is buying volatility. When the underlying is near a level at which a coupon would be payable, the issuer would prefer the stock to be highly volatile to provide a greater probable distance from the barrier.
If the underlying declines, the issuer still prefers high volatility to provide a greater probability of hitting the knock-in level when the issuer becomes long a put option.
This long volatility exposure is hedged by issuers selling volatility. The indeterminate maturity, however, makes this hedge highly dynamic and pressures the term structure and skew when flows are significant.
* As the underlying rallies, an auto-call becomes more probable and the expected maturity of the product shortens. Longer dated vega hedges are bought back to sell more shorter dated vega hedges. The opposite applies for a decline below the strike.
* Similar to the effect on term structures, as the probability of auto-call increases, the demand from issuers increases for upside calls to hedge the large coupon. Similarly, as the probability of the knock-in put being triggered increases, the issuer would look to sell downside puts in advance. Both of these effects serve to pressure the volatility skew.
Gamma wells are an interesting effect experienced by issuers of auto-callables. These are created when one or more institutions hold large derivative positions and the required hedging activity on the underlying asset begin to influence its price movements.
Long gamma means delta hedging pressure tends to dampen volatility and sometimes supports the share price near or above the strike price near maturity.
At the knock-in level and especially close to an auto-call date or expiry, the gamma becomes large, which means dealers buy more and more of the underlying on the way down, providing support for the market.
The caveat is that at some point, the issuer will consider the release of their delta position to be more important than delta hedging profits and gamma becomes irrelevant because the issuing bank sells its hedge. In practice, few dealers will fully hedge themselves when the market approaches the barrier and will often sell off the delta before the barrier is breached in an attempt to achieve the anticipated correct hedge ratio.
Auto-callables are significant, but vary widely in maturity and strikes making gamma wells at any particular underlying level or on any particular date relatively small compared to other market forces at this stage, in our view. At the current low implied volatility levels, even auto-callables are hard to price with significant yield. Demand may subside if volatility remains low and bond yields rise further.
This week's Learning Curve was written byGerry FowlerCFA andLamia Outgenza, equity derivative strategists atCitigroupin London.