I Paid 0.9% Over NAV for a Bond ETF in the Loudest Rate-Fear Week of the Year — $2,070 of Premium Gone in Nine Days

page four of a downloadable PDF

The fill confirmation was sitting in my inbox at 12:40 in the afternoon, on a Tuesday in March, and I remember the specific feeling of having done something smart. The trade looked right on every screen I checked that day. Two weeks later I was on the fund's own website, at my kitchen counter, staring at a premium history chart that showed exactly how I had been separated from $2,070.

I had snagged a bond ETF during the single loudest interest-rate panic week of the year, at 0.9 percent above the value of the bonds inside it. The premium. Nobody warned me about the premium, cuz everyone was busy arguing about the Fed.

the number on page four of a PDF

Every ETF publishes, every day, two prices. The market price, what shares trade for on the exchange. And the net asset value, what the underlying bonds are actually worth, marked by the fund. The gap between them is the premium or discount, and for most sizable stock funds it hovers near zero cuz arbitrageurs flatten it instantly.

Bond funds are different, and March taught me why. The underlying bonds trade over the counter, in a dealer market that slows to a crawl during panics. Pricing gets stale, the market price moves on fear in real time, and the NAV moves on appraisals that arrive hours late. In a rate scare, the market price can run far ahead of the marks, and the premium balloons, and anyone who buys during the balloon is buying a claim on bonds at a markup the bonds themselves do not support.

The fund's own historical premium chart, which took me two weeks to find on page four of a downloadable PDF, showed the premium at 0.92 percent the morning I snagged. Nine days later, after the Fed's testimony calmed the tape, the premium sat at 0.04 percent. The bonds inside barely moved all nine days. My market price fell almost the whole gap. That is the $2,070, detailed by a line on a chart I did not know existed.

why the panic prices the wrapper, not the bonds

The mechanics deserve one honest paragraph, because they will happen again. When rate fear spikes, sellers want out of bond ETFs today. The creation and redemption machinery let's hefty institutions deliver or absorb that flow, but during the scramble, the visible market price is set by the most frightened marginal trader, while the NAV is set by dealer quotes on bonds that may not have traded all afternoon. Fear lives in the market price. Patience lives in the NAV. Buy at the peak of fear and you pay fear's price for patience's assets, then watch the two converge when the shouting stops.

The convergence is not optional. It is enforced daily by arbitrage. Which means the premium I paid was never a fee I might get back through holding. It was a transfer, executed at the moment of maximum crowd distress, from me to whoever sold me those shares.

checking myself before the purchase

To be fair to March-me, I did diligence. I read duration commentary, I compared yields against Treasuries, I even checked the expense ratio and the spread history. I did everything except ask the one question that mattered: what is the market price relative to the value of the holdings right now? The premium data was free, public, updated daily, one click past every dashboard I actually used.

the porch conversation in June

My brother-in-law, a fixed income trader for twenty years, listened to the whole story on a porch in June and asked whether I had looked at the premium. When I said no, he did not laugh, which was worse. He said most retail investors never do, and that rate-fear weeks are precisely when the gap between the two numbers pays his colleagues' bonuses.

the pattern I now check before every fund purchase

Since March, my pre-purchase checklist has one line above all others, and it applies to anything trading on an exchange with an underlying portfolio. Compare the market price to the NAV on the day you are buying, not on average, not historically, today. If the gap is wider than the fund's own normal range, and you are buying at the wide end during a panic, you are not buying the asset. You are buying the crowd's mood, and the crowd's mood mean-reverts without you.

the other direction, in May

The same logic runs in reverse for discounts, and I have since made one purchase I am quietly proud of, a municipal bond fund trading at a 1.6 percent discount during a different scare in May. Same discipline, other direction. The discount closed within three weeks. Nobody wrote about it anywhere, because the crowd never celebrates the times its fear mispriced something.

what the whole thing cost, honestly

The $2,070 is the honest number, price drop from premium collapse, on a $230,000 position, in nine days, with the underlying bond prices essentially unchanged. The position itself recovered over the following months as rates settled, and I still own a version of it, so this is tuition rather than catastrophe. The yield was fine. The duration was what I intended. Every analytical choice I made was defensible except the timing of the entry, which was made at the exact second the wrapper was most expensive relative to its contents.

There is a version of this lesson where someone loses the same $2,070 to a bad stock pick and blames the company. I cannot even do that. The bonds did nothing. The index inside the fund did nothing. The gap between two numbers on page four of a PDF did everything, and it did it in nine days, in the open, published daily, in a file I never opened.

I have it bookmarked now. The chart takes thirty seconds to check, and thirty seconds of checking would have moved my Tuesday purchase nine days forward, into the week the premium had already collapsed, at almost exactly the same bond prices.

Thirty seconds. That is the entire distance between my March self and the version of me writing this, and the difference in net worth between those two men is $2,070, plus one chart bookmark, plus the slight permanent satisfaction of knowing exactly what the crowd charged me for its mood.