#104 | Proprietary Indices – Part 3

Executive Summary

Volatility control is one of the most common and essential, but least understood, parts of the proprietary index puzzle. By shifting constantly allocations between a higher-volatility proprietary index and low or zero volatility position in cash or fixed income, volatility control can deliver consistent option pricing and strong return optics (100%+ participation with no cap) without completely sacrificing the return edge of the proprietary index. But there’s a dark side – volatility control is quite complex. The return for the client at the end of the year is the result of potentially hundreds of different allocations throughout the year, making the final performance virtually untraceable. As is so common these days, the tradeoff for volatility control is complexity for performance. There are clear incentives for the agent, insurance company and the bank – but the ultimate benefits for the client are a little less clear.

Volatility control, in its simplest form, is most easily understood in the context of an index with two components – the S&P 500 and a cash account with zero yield and zero risk. Let’s assume that this index has a one-year volatility target of 10%. When the S&P 500 is trading at its average volatility of roughly 15%, a 10% overall volatility target for the index would mean an allocation of 67% in the S&P 500 and 33% in the cash account. But if volatility in the S&P 500 spikes to 25%, then the allocation shifts to 25% in the S&P 500 and 75% in cash in order to maintain the 10% volatility target. Conversely, if volatility in the S&P 500 drops to 10% or lower, then the exposure to the S&P 500 is 100% or higher, if leverage is required to meet the overall volatility target. How often the index is reshuffled to maintain the volatility target, which underlying assets or indices are included in the overall proprietary index and the ruleset for shifting the allocation are some of the ways in which banks differentiate their volatility controlled proprietary index offerings. But at the end of the day, this simplified model is indicative of how the typical volatility control mechanism works.

You can see the inherent attraction in the story. Volatility is volatile, after all, and volatility tends to spike when markets drop. That’s exactly the moment when an investor would like less exposure to the market and more exposure to a stable alternative like cash. The converse is true when markets are consistently rising. If you look at the charts of how these types of indices perform against the full-equity baseline index, you’ll see that the volatility-controlled version loses less in market downturns but only gains slightly less in market upturns. Who wouldn’t want that combination of returns?

But there’s a lot more to the attraction of volatility control than just the argument outlined above. Volatility control opens up a whole host of other benefits that are almost unavoidably intoxicating for life insurers building indexed insurance products – not the least of which is what it can do for the optics of their products. Let’s go back to the basic volatility-controlled index outlined above that allocated between the S&P 500 and a cash account in order to maintain a 10% volatility target. The cap for this account with a 2.5% option budget would be roughly 6.25%. The cap for the S&P 500 account, assuming historical volatility of 15%, would be 6%. But what if the volatility-controlled index had a 5% volatility target? Well, that changes the whole conversation. At 5% volatility, the price of an at-the-money call option is right at 2.5%. In other words, that account wouldn’t have a cap. It would be able to offer 100% participation with no cap on interest earnings. The optics of the product now totally change because it can consistently offer no downside risk and unlimited upside. The virtue of volatility control is not just that it can offer 100%+ participation with no cap but also that it can consistently deliver that level of participation. If volatility is consistent, then so is the price of the uncapped option [see Technical note at the end of the article]. A company can basically put all of its chips into an uncapped story without having to worry about a spike in market volatility wrecking it – and that’s exactly what a good number of carriers have done, especially new entrants like Nationwide that are now dominating the FIA space. It doesn’t take much imagination to see why this concept has caught fire in the Fixed Indexed Annuities.

And it doesn’t stop there. Remember that one of the core concepts behind the proprietary index is that it must (supposedly) deliver high risk-adjusted returns so that those benefits flow through the filter of the option pricing mechanism and to the client. Volatility control dilutes but doesn’t change the risk-adjusted benefits of the proprietary index. The bank can slap a volatility control mechanism onto a proprietary index and show 100%+ participation and no cap on the upside. The back-tested results obviously will look phenomenal. Volatility control and proprietary indices are a match made in heaven, especially when you consider that any proprietary index can fit nicely into an indexed insurance product’s option budget courtesy of volatility control. No matter what the volatility characteristics of the underlying index, the volatility control mechanism allows it to show 100%+ participation and no cap – nevermind the fact that achieving the volatility target means that the volatility-controlled index might be 10% in the underlying index and 90% in cash. As long as that 10% can show high risk-adjusted returns, that strategy will illustrate quite well. Volatility control opens the door to a whole host of indices that would otherwise not fit in an indexed insurance product.

[ Technical –  There’s another reason why banks love putting volatility control on proprietary indices that has nothing to do with client benefits. Banks make money on proprietary indices both by charging a fee for the index itself and by selling the derivatives that the insurance companies use to get exposure to the index. The problem with selling derivatives on a proprietary index is that there’s no market for pricing the derivatives. When a bank sells derivatives on an index like the S&P 500, it can reference the “volatility surface,” which is essentially the market pricing for the derivatives across different tenors and strikes. There is no volatility surface for a proprietary index because it doesn’t have a liquid options market. One solution would be for the bank to price a significant spread into the options on the proprietary index to make sure they cover the margin for error. But a much easier solution is to slap a volatility control mechanism onto the index, which essentially clamps down on the moving part of the pricing equation. By locking the volatility target, the bank can just price the option at the volatility target. It creates market pricing without a market, if you will. And without it, a lot of these proprietary indices wouldn’t price well or couldn’t even be traded. ]

Volatility control sounds great, doesn’t it? Magical even? As always, there are some wrinkles to the story, chief of which being the incredible complexity that it introduces into what might otherwise be a pretty simple structure. The moment you step into the world of volatility control, you step out of the world that your client can read about in the papers. Volatility control is its own animal. In my experience, many practitioners miss this distinction. They think that buying a S&P 500 Daily Risk Control 2 index is near-as-makes-no-difference to buying the S&P 500, but it’s not. It’s buying an algorithm that allocates between the S&P 500, Treasury futures and cash. And depending on how exactly volatility manifests itself in relation to the returns it delivers, that algorithm could work out well or not. You’ll never really know, actually. If you think I’m overplaying the complexity of the allocations and how often they change, just take a look at the chart on page 4 of this brochure. Every volatility-controlled index has a chart that looks like this one. Pick a year, any year, and you’ll be able to see huge allocation swings within the year. That’s just how volatility control works.

Consequently, it is going to be extremely difficult, if not impossible, to tell your client why their allocation between the sleeves within the volatility-controlled index is what it is at any given time. There are just too many factors and moving parts to the allocation strategies, particularly in some of the more complex indices like JP Morgan’s blockbuster Mozaic index. And it will therefore be extremely difficult, if not impossible, to explain to your client even why they got the credit they did at the end of the year because the final performance will be an amalgamation of potentially hundreds of shifts in the allocation throughout the year. Think this sort of thing doesn’t matter? Witness what happened in the Variable Annuity space when virtually every volatility controlled mutual fund managed to underperform for a few years and very few people, if any, could explain it to their clients. There was practically an uprising. Advisors were beyond furious. Ironically, those advisors started to flock to simpler products like Fixed Indexed Annuities, which in turn seduced them with volatility-controlled indices with 100%+ participation rates and no caps.

[ Technical – There’s a bit of a history to volatility control as well that should be mentioned. If the pitch for volatility control sounds eerily familiar, that’s because it’s the same basic idea as Portfolio Insurance, which was all the rage in the 1980’s and is commonly credited for the crash of 1987. But whereas Portfolio Insurance was built around portfolios of invested assets, volatility-controlled indices are non-investable and can only be accessed through derivatives, as discussed in the first post of this series. Don’t think, though, that the same challenges don’t apply. Portfolio Insurance was undone by a lack of liquidity in the futures market. When the algorithms called for mass selling of futures, it turned out that everyone’s algorithms called for the same thing at the same time and that kicked off the death spiral. In volatility-controlled indices, the switch from tracking the equity index to tracking a cash index is frictionless in the index, but hedging it is not. All banks dynamically hedge their option positions through other derivatives. To the extent that the composition of the index changes due to the volatility control rules, so should the composition of exposure to the index components through the other derivatives. Like Portfolio Insurance, hedging a volatility-controlled index means an assumption of boundless liquidity. There is a veritable graveyard of trading strategies that relied on liquidity for their execution, which was plentiful when times were good and thin when times were bad. We’ll see, I suppose. ]

But, for now, the appeal of volatility control for the agents, insurance companies and banks is undeniable. Agents love telling their clients that they have no risk and can get unlimited upside potential on a volatility-controlled index created by a firm with an unimpeachable name like JP Morgan. Insurance companies love volatility control because it creates stable caps and allows them to differentiate their offerings from their peers without increasing their risk or cost. And banks love volatility control because it opens the sandbox to a host of index constructions that would otherwise not work or only work at a prohibitive risk or cost, all while earning fees for the derivatives and on the index itself. It’s a really good gig. And it’s only getting better. Volatility control itself is even evolving. Remember how I drew a distinction in the previous posts about the difference between Blended indices and volatility control? That distinction is rapidly vanishing. If historical fixed income returns look great in a backtest for Blended indices, then why would a volatility control mechanism not allocate to fixed income rather than cash? Wouldn’t it be better to allocate to fixed income when yields are healthy and prices are stable and only allocate to cash when things really go haywire? Yes, of course it would be better. Just take a look at how S&P has updated their benchmark S&P 500 Daily Risk Control Indices to version 2, which now includes a fixed income component based on 10 year Treasury futures. No surprise, version 2 handedly outperforms version 1 in backtested returns. Using these new types of indices, volatility control in and of itself can generate high backtested returns. It’s a brave new world.

You might have noticed that I left out a not-inconsequential part of the value chain in the list of parties who benefit from volatility-controlled indices – the client. And in the last post of this series, we’ll explore the real risk and return potential of proprietary indices and how to appropriately position them as they start to expand into Indexed UL products.

[Technical – Volatility control takes care of the major pricing component of an uncapped option but the price of the option is still exposed to interest rate risk. If interest rates increase, then so will the price of the option and that will put pressure on the story of uncapped participation in the index. The solution that banks are increasingly starting to build into their indices is to make them “excess return” only. In its simplest form, excess return indices simply subtract the risk-free rate from the returns in the index. This eliminates the interest rate component of the returns and the option pricing. As a result, the participation level is no longer exposed directly to market interest rates, even though the earned rate and option budget at the insurer obviously still can impact the participation level. When interest rates were low, the idea of rising rates hurting FIA and IUL products was pretty foreign, but the front end of the yield curve has been creeping up and so have options prices. As a result, a lot of renewal rates on FIA products are taking a beating. Although I don’t particularly like excess return indices because they add another layer of complexity with no real benefit to clients, I can see why some insurers will start using them. If they don’t, they’re going to end up having to explain to clients why interest rates are going up and their caps are going down. That’s never a fun conversation, no matter how justified the changes are.