Synthetic structured credit, or synthetic tranches, may not have been the cause of the financial crisis but it was one of the ways that sophisticated financial participants convinced themselves that they weren’t really taking huge, irresponsible risks. So when Citi analysts Abel Elizalde and Aritra Banerjee predict that the synthetic tranche market is going to make a comeback it’s bound to make people uncomfortable, and the report makes it clear why so many people are nervous about returning to such pro-cyclical instruments.

“We’ve written at length recently about how investors will be forced to take synthetic levered risk – partly via indices but mostly via tranches. Our 2015 ‘outlook’ had a very clear title: Keep calm and lever up (synthetics),” they write.

implied vol v spread synthetic tranches

Tranche prices depend on four factors, but they can’t all be reliably estimated

As a quick review, if you want to figure out what a tranche is worth you need to know four things: the characteristics of the security being offered (maturity and subordination), the likelihood of default in the underlying credit, the expected recovery rate, and default correlations. The characteristics are given, and recovery rates can be estimated, but the other two factors can get investors into trouble pretty rapidly.

In the case of the financial crisis, the probability of default was thought to be low because housing prices would keep going up and up until the end of time – which really is hardly an exaggeration if you remember the sentiment before the crisis. The argument that access to cheap credit will keep even unhealthy companies from going bust could easily translate as underpricing default risk in synthetic tranches.

[drizzle]Figuring out default correlations is even less reliable. Basically you just have to use current market prices and then figure out the implied correlation.

“Don’t we do the same when pricing options? We derive the ‘implied volatility’ from the market price of the option using the Black-Scholes model … the wrong parameter in the wrong model which gives the right price,” Elizalde wrote last May (n.b. ellipsis in the original).

But Elizalde points out in that paper that implied default correlations vary by tranche, which would be impossible if they reflected default correlations in the real world. You might even conclude that investors are just choosing the yields that meets their investment needs without thinking too hard about the risk.

Even if tranches are a bad idea, it doesn’t mean they won’t become popular

Elizalde and Banarjee understand how bad this all sounds, but if you take their argument purely as a forecast instead of a recommendation it makes a lot of sense.

“The tranche market isn’t one which triggers fond memories for many people. We are not expecting investors to go long via tranches because they want to, but because they need to,” they write.

Synthetic structured credit gets more popular as spreads tighten (check, thanks to globally loose monetary policy) and mark-to-market volatility is low (falling again now that the Greek debt situation has been successfully papered over). Elizalde has called tranches the most pro-cyclical product there is, so if you really need to lever up and the conditions are right then demand will start to grow. The combination of implied vol and credit spreads is getting closer to 2006 levels when synthetic tranches were last popular. If statements like that set off alarm bells you’re not alone, but you can’t deny that many investors are reaching further and further for yield.