Hedge cost as a permanent drag versus running unhedged and eating the variance

Question

I've been working through the interest rate hedging disclosures on a handful of agency-focused mREITs and the range of approaches is wider than I expected for what's supposedly a commodity business.

One group runs a heavy hedge book, swaps and treasury futures against most of the repo funding. The economics are that you give up basis points every quarter to buy stability in net interest income and book value. In a flat environment that cost is pure drag and it shows up in a lower dividend than the unhedged peer. The defense is that the drag is the price of not being a forced seller when the curve moves, and that surviving is the whole business when you're funding long assets with short borrowing.

The other group hedges lightly and takes the position that hedges are imperfect anyway. Swaps hedge parallel moves reasonably well and do very little for a steepener or a twist, so you pay full cost for partial protection. That group argues shareholders bought a spread business and should get the spread, and that management pretending it can neutralize rate risk is selling comfort rather than delivering it.

What makes this hard to resolve is that both positions look right in different periods, and the disclosure is thin enough that you often can't tell what the hedge actually accomplished. The sensitivity tables show parallel shifts of fifty and a hundred basis points, which is the scenario the hedge is best at. The scenario that hurts is the one nobody tabulates.

So where do you come down. Pay the drag, or take the variance and demand a higher yield for it.

chriswalton8 · Answer

Anchor, the part I'm stuck on is that the choice you're describing isn't really available to a shareholder.

You can pick an issuer whose current hedge posture you like, but hedge policy isn't a covenant. Management can be seventy percent hedged this quarter and thirty next, and you find out after the fact in a filing. So what you're actually choosing is a management team's judgment about when to hedge, and that's a much harder thing to evaluate than a static policy.

Which pushes me toward the heavy hedge side, for a reason that has nothing to do with the math. A team that pays the drag consistently is telling you something about how it thinks about tail risk, and I can check whether the stated posture matches the disclosed positions quarter after quarter. A lightly hedged team might be right about swaps being imperfect and might also just be reaching for yield with a good story attached.

What I can't judge is whether the drag I'm seeing is a reasonable price or an expensive one. Comparing hedge cost across issuers seems close to impossible given how differently they present it.

luis_carrera · Answer

Sable's point about policy not being fixed is the one that broke my spreadsheet.

I'd built a model comparing hedge ratios against realized book value volatility over sixteen quarters, thinking I'd find that the heavier hedgers had smoother books. The relationship was much weaker than I expected, and when I looked at why, it was because the hedge ratios themselves moved around so much within the period that the average I was using described almost nobody's actual position.

So I'm voting for the higher yield with variance, and my reason is close to giving up on the analysis. If I can't distinguish good hedging from lucky hedging in the disclosure, then paying the drag is paying for something I can't verify. I'd rather hold the unhedged version where at least I know what I own, size it smaller, and treat the variance as the thing I signed up for.

Anchor, on the sensitivity tables only showing parallel moves, has anyone here found an issuer that tabulates a steepener. I've looked and every table I've pulled runs the same parallel scenarios.

Hedge cost as a permanent drag versus running unhedged and eating the variance

2 replies