Now Is Not The Time To Lean Into The Illiquidity Premium In Credit, One Bank Warns

If you’re looking to lean into the illiquidity premium in an environment where yield is becoming an endangered species, this is either the best of times of the worst of times, depending on your penchant for risk.

One one hand, the spread pickup for rolling the dice on illiquid credits is rising both in high grade and junk, with the latter looking to make a run at post-2016 peaks.

The premium is near two-year highs for both IG and HY, according to portfolios constructed by Goldman, who sorted bonds based on their price impact using TRACE data.

(Goldman)

That’s the good news. The bad news is, chasing that is likely to leave you in the lurch if volatility picks up further in an environment where “tweet risk” (so to speak) is as high as it’s ever been.

In the following chart, Goldman illustrates the rates-hedged performance (cumulative excess return) on a long-short strategy involving illiquid versus liquid bonds.

(Goldman)

Clearly, illiquid high yield has not performed well and the reasons are straightforward.

“We think weak growth sentiment, lingering policy uncertainty, and relatively expensive valuations leave ample scope for further underperformance of illiquid bonds”, Goldman cautions.

The bank also notes that underperformance by illiquid issues could well be exacerbated by “the more fragile post-crisis market microstructure”.

How comfortable are you in illiquid credit, when corporate balance sheets are overleveraged, the cycle may be turning and, thanks to the onerous post-crisis regulatory regime, the street is constrained in its capacity to act as middleman and warehouse risk?

(Deutsche Bank)

But hey, none of this need apply, because after all, you can just load up on high yield and emerging market bond ETFs. Everybody knows that transforming an illiquid pool of underlying assets into units that are perfectly liquid all day, everyday is as simple as securitizing them into a BlackRock-sponsored retail product, right?

When has that kind of thing ever gone wrong?