“Price”, ETFs, and Bond Market Liquidity

Multiple events last year reminded us that “price” is a nebulous concept. The most well-publicized price disagreement came in September of 2019 when the public market balked at the price the private market (ok, mainly Softbank’s CEO Masayoshi Son) set for WeWork. But public market participants also disagreed on WeWork’s value. From $58 a share pre-IPO, the price recorded by mutual fund owners dropped post-IPO to $25.8. No, wait…Fidelity had them at $35.33, Vanguard at $45.90, T. Rowe Price at $17.03 share…yes, that’s my point.

The recent pricing of the Saudi Aramco IPO is another fun example. Unable to convince international institutional investors to price 5% of the company at $1.7 trillion dollars, the Saudi government shrank the number of shares down to 1.5%, limited the offering to the Saudi exchange, and enticed (and threatened?) local retail buyers with added options. Was the resulting $8.53 a share valuing the company at $1.7 trillion correct? Maybe it was too low as 3 days later, the stock traded up to a $2 trillion valuation likely making some hedge fund managers happy as opined by Bloomberg’s Matt Levine.

If these examples seem interesting but disconnected from your portfolio, read on. Questionable prices lurk in many fund portfolios, especially those invested in bonds and loans. I will give a couple of both historic and current examples, including within the leveraged loan market, media’s favorite liquidity boogie man. Soothing your liquidity fears demands some specifics on how bond Exchange Traded Funds (ETFs…the source for much of the angst) work. First, though, some more thoughts on “price”.

Tesla’s Price, For Example

A liquid and transparent market like the stock market provides a price that we can all either agree on or bet against. Take Tesla. As reported by FT’s Alphaville, Morgan Stanley equity analyst Adam Jonas believed in December that the stock was worth $250, 25% or so below its then-current price. But as shown below, he’s acknowledged it could go to $500 just as likely as it could go to $10. He recognized that his view of the correct price could be very wrong. In fact, he was very wrong. The stock traded well north of even $500 by early this year and in February came close to doubling that number!

But the then $333 per share market price wasn’t necessarily wrong. That single clearing price at the time reflected Adam’s and thousands of other views. We can argue all day on whether that price was biased by a combination of behavioral biases and risk attributes, but for sanity’s sake, let’s agree that this equilibrium price was generally the best guess of value then, despite it being grossly wrong with hindsight. With a little help from the regulators and a lot of help from technology, most published bond prices are also good momentary indications of value.

How Most Bonds are Priced

Bonds lack a centralized market, but most corporate and municipal issues, all treasuries and all agency mortgage-backed bonds (MBS) still benefit from regulated price transparency. FINRA demands that broker-dealers report their trades through TRACE (Trade Reporting and Compliance Engine). If your investment advisor has bought you a bond and you are curious about how the price compares to what others have paid, check out FINRA’s website here. 

Trades that involve private placements like leverage loans and those backed by seasoned private-label MBS don’t share that same requirement but still have a degree of price transparency. Private Label MBS transactions are reported to pricing services like IDC, Reuters, and Bloomberg, and the leveraged loan industry group LSTA likewise tracks and disseminates member secondary trading data. But because many of these bonds and loans are not widely held and trade infrequently, their prices can still be the result of a bit of guesswork.  Sometimes a lot of guesswork.

A Case Study: Aviron Capital’s Loan “Price”

Take the closed-end fund, The BlackRock Multi-Sector Income Trust (BIT), which as the name implies is a fund that holds a hodge-podge of bonds and loans. As of October 2018, included in this hodge-podge was a $63+mm loan to Aviron Capital with a generous LIBOR+500 coupon.  That was at the time nearly 9% of the fund, at least as valued by BlackRock given the loan was classified as a Level 3 asset, an asset designated to have little, if any, price transparency.

After a bit of digging, BIT seems to have largely financed Aviron Capital (owner of Aviron Pictures) which in turn distributes big-name Hollywood films like Serenity starring Matthew McConaughey and Anne Hathaway. It sounds like fun…at least for the Portfolio Managers. If you haven’t heard about Serenity, you’re not alone. It was a bust.

By April of 2019, Serenity’s dud status was secure.  BIT’s semi-annual report shows an additional 10mm notional value to the holding but an implied “price” closer to 75 cents on the dollar.

By July, the “price” dropped further, to roughly 70% of the dollar, but still nearly 6.9% of the portfolio (as reported by Bloomberg).

By September, according to the BlackRock’s Fact Sheet, the price dropped even further. Assuming no magical buyer emerged, the position by then represented only 3.78% of the fund’s $688.9mm in net assets implying a total value of $26mm or 35.5 cents on the dollar.

Sadly, it didn’t stop there. A drop in the fund’s NAV by 1.9% on Oct 31^st and a reported holding percent of 2.28% implies roughly a $15.38mm value or roughly 21 cents on the dollar — yikes!

BIT’s NAV took another 1.9% hit on December 10^th, and a final 1.4% hit on December 18th. Not only did BlackRock completely write down the position by year-end, they wrote it off retroactively to zero as of the October 31st Annual Report and adjusted the NAV down from $17.93 to $17.28.

Were any of those previous “prices” as calculated by BlackRock accurate? My point is that without a liquid and transparent market, who knows? ⁽¹⁾

Virgin Trains USA’s Bond “Price”

What other potential landmines may be lurking out there? I wish I knew. And I wish I had some idea of the feasibility of a passenger rail car from West Palm Beach to Orlando, but somehow Nuveen did. When Virgin Trains USA (owned by Softbank’s Fortress with a “Sprinkle of Virgin Magic” by Richard Branson), priced $1.75 billion of non-rated tax-exempt bonds, Nuveen reportedly bought well over half of the offering.

Does having a large lead, especially a well-respected muni behemouth like Nuveen, owning such a large percentage of an issue create a problem? Arguably not, but who else is going to provide them much liquidity on close to a billion dollars in non-rated bonds should Nuveen change its mind?

Granted, millions of dollars of Virgin Train USA bonds trade every month, but other buyers seemed slow to emerge despite a rallying high yield muni market. Maybe it was the escalating death rate along the portion of the line now currently operating. Or perhaps buyers found worrying that portion of the line’s stagnant ridership growth. Or maybe even the legal battle on whether passenger railroads can even issue tax-exempt private activity bonds designated for highways is giving them pause ⁽²⁾. It’s early, so hard to tell. The point is that the price for 1,000 bonds ($1mm) is likely very different than 752,460 and likely would have been very different without Nuveen’s belief in the project.

Size Matters, at Least to Private Label MBS Traders

Size matters in other markets, too. But sometimes the problem is that the trade is too small, especially if it takes a bit of work to figure out how the cash flows work. What trader, for instance, is going to spend much time trying to figure out the value of bonds like those private-label legacy MBS “odd-lots” below. As an inducement, the offer indicates an investor can buy them and immediately show a profit when IDC prices them higher as a full position.

I’m not picking on IDC. As an ex-salesman for bonds like the above, I can attest to how much guesswork and imprecision often goes into valuing idiosyncratic bonds. In fact, one of my roles throughout the 1990s — before firms like IDC became ubiquitous — was to fax over daily prices from my traders to my mutual fund clients. That task wasn’t a challenge most of the time, but when liquidity evaporated, sparks flew!

Several examples of these sparks and the resulting fires: in 1994, The Piper Jaffrey Institutional Government Income Fund dropped 25% as the liquidity and value of a slew of mortgage derivative evaporated because of higher than expected interest rates and resulting in lower prepay assumptions. I can attest that the assistant portfolio manager would call up in tears pleading for more reasonable prices. They were later fined when they instead simply adjusted the prices themselves. In early 2007 as the subprime mortgage-backed securities market began to crumble, James Kelsoe’s Jr., a heralded fixed income portfolio manager at Morgan Keegan, made 262 “price adjustments” according to the SEC while trying to prop up the fund’s price.

But most disputes about the correct price are reasonable. Figuring out the correct “price” for illiquid bonds, be it for reasons of complexity, size, opaqueness, or market disruptions, makes determining the Net Asset Value (NAV) for a fund that holds these bonds tricky at best. So how nervous am I about the potential mispricing during periods of market stress of rapidly growing, complex leveraged loans, often issued by opaque small private companies, and traded without the benefit of published prices in a growing array of retail funds? Joining a full chorus of anxious regulators, journalists, analysts, politicians and pundits, anyone unaware of the potential for leverage loan prices to gap down in the face of higher default is naïve. Where I perhaps differ is my faith in leverage loan ETFs to provide the sector with needed price transparency, not so much in the reported NAV as with the actual trading price of the ETF.

Fixed Income ETFs like those for Leverage Loans Provide Price Transparency.

After growing over 20% in 2018 and over 15% on average since 2000, faster than other credit markets, the $1.1 trillion dollar leverage loan market took a pause in 2019 and may end up the year slowing by over 25%.

In fact, the pause started in late 2018, led by large outflows from mutual funds and ETFs.

In December alone, $15bn in net outflows representing nearly 10% of the retail fund holdings occurred and those outflows continued into 2019.

How was liquidity? According to a study by the LSTA, loan trading volumes actually increased and prices dropped a pretty modest 2.9%. With heavy trading, leveraged loan ETFs dropped even further than their underlying loans, a fact that deserves a deeper dive into the unique mechanics of fixed income ETFs.

Fixed Income ETF Mechanics

Like their equity ETF brothers, fixed income ETFs employ authorized participants (APs) to make markets in their shares. To enhance liquidity, the APs can deliver bonds into the ETF manager in exchange for newly issued ETF shares. The profit opportunity to buy the bonds at a lower overall price than the value of the new shares motivates the AP to keep the ETF shares from rising too high relative to the underlying bonds. If the ETF shares trade lower in aggregate than the bonds, the AP can profit by redeeming shares for bonds. This creation/redemption ability theoretically keeps ETF share prices trading close to their reported NAVs. ⁽³⁾

That’s all fine in theory, but in fact, only a small fraction of the bonds held in an ETF are available to the AP for delivery at any one time.  In addition, as we’ve been discussing, the funds also may hold illiquid bonds and loans with little price transparency. The AP may not want to receive those bonds in a volatile market. During periods of market stress, calculating an accurate NAV is thus no small task. ETFs handle these challenges in different ways, but primarily they use “custom baskets” of bonds and allow APs to use cash (net execution costs) when creating and redeeming shares.

Custom Baskets and New SEC Rules

Instead of providing a pro-rata portion of every single bond, most of the more established fixed income ETF providers were given exemptions to create customized baskets of bonds that could be used to recreate or redeem shares. The SEC’s new rules, finalized in September, extended that right to all firms willing to establish basic procedures. More, these baskets can now be modified over the course of the day. This facilitates the ability to negotiate large credit portfolio trades with institutional counterparties. It also allows for the flexibility to exclude illiquid holdings.

Cash Creates and Redeems

Even with custom baskets, many portfolio managers will continue to allow APs to use cash instead of bonds when creating additional ETF shares. In fact, cash is often preferred by the ETF portfolio managers (PMs), especially if they want to take advantage of new issues or what they deem to be cheap offerings the AP doesn’t hold. For many ultrashort fixed income ETFs, only cash can be used to create or redeem ETF shares.

Seems like this should allow APs to keep the ETF trading very close to the NAV and most often they do. But not always as we saw last December.

SRLN’s “Price”

SRLN is SPDR’s leveraged loan ETF called the Blackstone/GSO Senior Loan ETF which is actively managed by GSO. I pick on SRLN because I like them (and own it), but when the ETF price dropped to a 1% discount to its NAV, I did pause. In fact, SRLN was one of many fixed-income ETFs that traded briefly at more than a 1% discount to their NAVs. Obviously, there are additional charges and costs, especially in illiquid markets, that APs incur.

Additional Charges and Costs for APs

For starters, the Authorized Participant (AP) only knows the cash amount of the shares at the end of the day when the NAV is calculated. In a volatile, illiquid market (like Christmas week of last year) making too fine of an arbitrage calculation without yet knowing the ETF NAV is risky. APs call this risk “slippage”. Further muddying the waters, ETF PMs also try and protect their shareholders by charging back execution costs on cash redemptions. This charge basically adjusts the cash proceeds going to the AP to reflect the actual selling prices achieved to raise the cash.⁽⁴⁾ SRLN’s APs appear to have had only so much tolerance to take on that execution risk during Christmas week.

Other times, the APs can seem to transact at inopportune times. For instance, an AP opted to create 2.7mm of SRLN shares on January 4^th even when the shares were trading below NAV. Why not just buy those shares in the market? Perhaps the AP found itself short shares after the tremendous buying the first few days of the year.  If so, in hindsight, accepting the lower NAV price versus market price was a prudent decision. The $45.5 per share NAV price the AP received was lower than the $45.60 the ETF shares traded at during the first half-hour of trading versus the next day.

Finally, even in the course of everyday business, the ETF will often charge a bid/ask when redeeming or creating shares for cash. This can vary from a couple of basis points (bps) to 20bps depending on the ETF and is subject to change.

Mutual Funds are Simpler, but for a Price

If this sounds too complicated and makes you want to sell your ETF shares for a mutual fund, consider who instead bears the cost of liquidity of a mutual fund: the shareholder for the benefit of the trader. The trader is allowed to trade in and out of a mutual fund at NAV and perhaps a modest ticket price. But the bid/ask of trading the individual loans doesn’t go away. They are absorbed into the returns of the fund. As we’ve shown, those costs can be very high when liquidations are high, markets are illiquid, and the NAV doesn’t match the prices received.

Even when the assets are on a central exchange, complications arise when they are in different time zones like shares in international funds. After years of being taken advantage of, mutual fund managers have learned to incorporate “fair-value pricing” when setting a 4 pm Eastern Time Zone NAV for international funds. What do some use for one of their inputs: ETF prices!

Conclusion: ETFs Bring more Price Transparency to Markets

Far from alchemy, the ability of ETFs to centralize price discovery creates more price transparency to otherwise opaque asset segments. That doesn’t imply that the idiosyncrasies of underlying securities disappear. “Price” will always depend on size and ownership concentration. More, when assumptions change, large gaps in price can occur. All you have to do is look at the history of MBS: billions of mortgage back loans now trade every day, but only as pools and bundles of loans. There is little liquidity in individual loans. And when default or prepay assumptions change, even those pooled securities can exhibit large price gaps. Piper Jaffrey and the Morgan Keegan funds in 1994 and 2007, respectively, are two prime examples.

But the brief decoupling of ETF prices from their NAVs in the illiquid markets of December of 2018 was a feature, not a bug. Those prices reflected a much truer cost of liquidity than their NAVs and rewarded investors who were willing and able to provide it. For those worried about schizophrenic retail holders of fixed income ETFs, lower prices should discourage spooked holders from selling and bring in value buyers. SRLN’s heaviest trading volume days all year by a wide margin were from December 20^th through the 26^th.

My own practical advice: next time a comparable ETF to your mutual fund trades at a significant discount to its NAV, considering swapping into the ETF from your mutual fund. The mid-market NAV “price” you are getting from your mutual fund doesn’t likely reflect the true cost of liquidity as is reflected in the price of the ETF.

My more general advice is to be thankful that ETFs are providing price transparency to once opaque markets, not fearful. More buyers and sellers are a good thing. In fact, large APs like Jane Street are replacing broker-dealers as providers of liquidity on individual bonds and now even hold larger bond inventories. Just don’t blame the ETF structure if your leverage loan fund drops 4% again…or 10%. That risk is hopefully now more transparent, too.