3 Overreactions To PIMCO’s Probe
The PIMCO investigation has turned a spotlight on bond pricing.
The PIMCO investigation has turned a spotlight on bond pricing.
The PIMCO investigation has turned a spotlight on bond pricing.
My piece yesterday talking about the SEC investigation of PIMCO’s pricing mechanics in the PIMCO Total Return ETF (BOND | B) raised more than a few eyebrows. It also generated some email traffic and broad comments, which I’ll be generous and call “heated.” Apparently, there are a lot of folks out there who feel very strongly one way or another about this. It seemed like answering a few questions here might be easier than writing 100 emails saying the same thing.
First off, I want to be clear: I don’t have any more information than anyone else about exactly what the SEC thinks PIMCO might have done, or whether PIMCO actually did whatever the regulators think it did. I was responding to the same news story everyone else read, which suggested that PIMCO was buying at a discount, then using a pricing service to value its bonds. I stand by my belief that both of these are not only standard practice, but pretty much the only option.
If in fact the SEC thinks PIMCO is doing something different and nefarious, well of course they should be held accountable. And by “they” I really mean the fund board, which is the watchdog on pricing policies.
With that out of the way, let’s look at a few common objections to what I wrote yesterday.
“Dave, BOND is going to tank, and the market’s never going to trust PIMCO’s prices again!”
Most of the comments along these lines came at me in the afternoon, when this “scary” chart was unfolding:
The blue line here is how BOND actually traded yesterday. The black line is the intraday net asset value being reported for BOND. Look at how it fell out of bed! Trading to such a crazy discount! The problem is scale. BOND traded to a discount of 0.11 percent. This is entirely within the bounds of PIMCO’s daily fluctuation from reported value. Just look at the last 10 days:
Down 11 basis points yesterday, but at a premium on Sept. 19. Discount on Sept. 18? And look, a premium on Sept. 17.
In all cases, enormously tight to NAV. Expect this pattern to continue. Why? Because the NAV isn’t just some made-up price used to calculate returns. It’s what PIMCO pays an authorized participant on a redemption.
But at any discount, the APs can just buy up shares of BOND in the open market, hand them back to PIMCO and collect the “overstated” fair value. Most likely, they’ll even get it in cash, because PIMCO generally prefers to do its own bond trades rather than in-kind.
“But Dave, you said bond ETFs are systematically understating their value, but look at (insert ETF!). It’s trading at a discount!”
It’s absolutely true that on any given day, any ETF is going to trade at a premium or discount to its reported NAV. That’s just how the market works. The question is why. In most cases, the single most important factor is flows. Take a look at the pattern of premiums and discounts inside BOND over the long haul:
The middle line here is the premium/discount. The bottom column chart is fund flows. Notice anything peculiar here?
During 2012 and the first quarter of 2013, BOND received enormous flows, and consistently traded at a premium. During most of the time since, BOND has received consistent outflows, and has traded at a discount. In late summer of this year, BOND had inflows again, and voila, traded to a premium. In most cases, all of the swings were within a dozen or so basis points.
This isn’t broken; this is exactly how the ETF creation/redemption process is supposed to work. Remember, investors don’t “put money into BOND.” They buy BOND shares on the open market. When enough investors do that at the same time, they force BOND to trade to a premium. The AP participants see that premium and pounce, manufacturing new shares to book the premium as a profit.
Consider how a less liquid ETF, the Market Vectors High Yield Muni ETF (HYD | C-74), functions. This is a super tricky and illiquid corner of the market. And it shows:
When investors wanted out of the fund back in 2013, it traded to significant discounts—almost 6 percent. Why so much?
Well, my suspicion would be that the AP looked at what they were going to receive if they agreed to accept a redemption basket from Market Vectors and didn’t like what they saw. If Market Vectors had said, “If you give us 50,000 shares tonight, we’ll just give you cash,” then they would have bought up shares in the open market long before the price dropped 5 percent, and booked that discount early.
That they didn’t means they saw something in the basket whose price they didn’t believe—actual high-yield municipal bonds. Since those things hardly trade, it’s anyone’s guess what they’re actually worth, and pricing services tend to react very slowly to changing market conditions.
Over time in the summer of 2013, the NAV of HYD—again, determined most likely by pricing-service data—did collapse, though slowly. When investors really wanted out, the actual price of HYD is what led the market.
In other words, HYD was price discovery.
“But Dave, where’s the persistent ‘sandbagging’ of portfolio values you’re talking about?”
The short answer is almost everywhere. No, it’s not present in the high-yield muni space when the market’s selling it off, because flows overwhelm it. Much more normal is something like what we see in the iShares iBoxx $ High Yield Corporate Bond ETF (HYG | B-74).
You can still see the impact of flows here. Look at what happened just a few months ago when there was a big sell-off in junk bonds: a rash of redemptions, and HYG trading to a 0.58 percent discount. But the norm here—the average day—is a consistent 0.22 percent mispricing of the NAV versus the ETF.
The market realizes that the HYG portfolio is being priced on the bid, and in the absence of flows, APs just let it trade at a little bit of a premium to account for this, because they know they can’t build the creation basket for fair value.
We see this in virtually every corner of the market. Even the $18 billion, 1-million-shares-a-day giant that is the iShares Core U.S. Aggregate Bond ETF (AGG | A-97) trades at a consistent premium of between 0.05 percent and 0.10 percent.
Finding A Fix—Good Luck
I like to think I’m a realist. But what’s wrong here isn’t the ETF structure or how people price their funds. With bond dealing a dying profession, and bonds themselves starting to price less like stocks and more like fine art every day, it’s unclear exactly what the folks looking for their pitchforks are actually expecting bond managers to do.
For a small, illiquid bond, there is no functioning market. The bond market is getting worse, not better. Talent and capital is leaving the business of providing efficient bond markets. The people left are either buyers (like PIMCO), or sellers (banks and issuers).
If you take a veteran bond dealer out for a beer, they generally glaze over and start looking like Roy Batty sitting on the roof at the end of “Blade Runner,” saying “I’ve seen things you people wouldn’t believe” before diving into a monologue about the death of Drexel Burnham. There’s a good chance he’s not in the business anymore, or is looking to get out.
I don’t have a solution for this. I don’t know how you manufacture a functioning market out of thin air. The “good old days” for bonds were not all that great, or “fair” either, and trying to recreate some high-tech version of the old human bond dealer isn’t just going to magically fix things. “More electronic trading”—a popular answer in my email inbox—doesn’t fix a thing if the bonds in question do not trade.
At the time this article was written, the author held no positions in the securities mentioned. You can reach Dave Nadig at [email protected], or on Twitter @DaveNadig.
