Two charts in a recent presentation by Citi’s credit product strategist, Matt King, tell the story of a structural fault line in the financial markets.
King illustrates how the assets of US mutual funds invested in investment grade and high-yield debt have trebled since 2008, from $300 to $900 billion, reflecting investors’ increasingly desperate search for interest income in low-rate environment.
But the infrastructure supporting bond trading has weakened. The banks that act as intermediaries between the bond market’s buyers and sellers have seen their holdings drop from $286 billion to $69 billion—by 76 percent—to levels last seen in 2002.
And King reminds us that bond prices are much more prone to jumps and “cliff effects” than share prices: as an example, note the 2006-08 behaviour of Lehman Brothers’ equity and of the company’s 2016 euro-denominated bond (illustrated in the chart below).
Lehman’s share price declined steadily over the two-year period leading up to the company’s insolvency as investors factored in its deteriorating prospects: but Lehman bondholders didn’t face up to the worst until right at the end, at which point bond prices went from 70 to near-zero in an instant.
This kind of behaviour places an automatic strain on the relationship between bond funds—which promise instant liquidity and continuous pricing—and the underlying bond asset class.
Meanwhile, the underlying liquidity of the bond market has deteriorated since Lehman went bust. Efforts at forcing bond trading into a central, exchange-like format haven’t worked. The market still relies on a dealer-centric, capital-intensive transaction model. And the dealers’ capacity to hold bonds has been steadily eroded since 2008 due to restrictions on proprietary trading and higher regulatory charges for inventory.
When an asset manager wants to buy or sell bonds he typically asks several dealers via a so-called request-for-quote (RFQ) process. Dealers and clients are aware of each other’s identity all the way from the initial request through to post-trade clearing and settlement.
Bond dealing is fragile and rests on trust (in particular that pre-trade disclosure of an asset manager’s intentions won’t be abused). Dealers are under no obligation to buy or sell bonds they don’t want and investors fear that prices could move against them.
Meanwhile most bonds see little, if any trading at all. Investors have little idea of the real price of individual bonds until they test the market. Dealers’ daily price “runs” (lists) are indicative, non-binding and often biased to reflect their trading preference. In bonds, there’s no deep, transparent, anonymised market of the type we find on most stock exchanges (equities are usually traded via a so-called “central limit order book” or CLOB).
Add this fundamental difference in market conduct between bonds and equities to the way many bonds are currently owned and there’s a particular problem, say Larry Tabb and Will Rhode, authors of a recent study (“US Corporate Bond Trading 2013: In Search Of a New Market Structure”—the study itself carries a hefty subscription fee, but you can read an interesting debate on the topic on the Tabb Group’s website, for which registration is required).
“The low levels of dealer capital would probably be fine if all bond inventories were held by individual investors, who are less likely to liquidate their holdings in the event of a rate rise,” say Tabb and Rhode.
“However, retail investors’ bond holdings are increasingly sitting in mutual funds and ETFs. Packaged fixed income products such as ETFs and mutual funds are perceived to be much more liquid and easy to sell than the bonds themselves. However, the truth is that redemptions will force bond managers to sell the underlying bonds….What concerns dealers and asset managers alike is the fact that an interest rate rise causes investors to liquidate fixed income funds…causing a flood of bonds to hit the market at a time when there is little dealer liquidity to absorb the shock.”
You won’t find the managers of corporate bond funds or ETFs advertising this risk, except indirectly; last year asset manager M&G acted to slow inflows into its corporate bond fund, though the firm explained this at the time as a policy to help the manager invest funds more easily, rather than as a precautionary measure against potential redemption pressures.
But we saw a dress rehearsal of chaotic bond fund redemptions in June this year when the Fed hinted it might raise rates. The ensuing market disruption then caused the US central bank to backtrack rapidly.
The risk of large-scale bond fund runs remains. “I know first-hand that not only the buy-side but the sell-side is scared,” says Larry Tabb.
Many buy-side firms have made efforts to overcome the structural imbalance in bond trading. There have been several initiatives to start trading networks involving the big investment firms. But so far these have all failed to reach critical mass. Asset managers are also reluctant to reclassify themselves as bond dealers, for fear of incurring heavier regulation. As a result, funds still rely largely on bank dealers to enter and exit positions.
And in its study Tabb reveals an impasse between the buy-side and sell-side when it comes to the future direction of corporate bond trading.
100 percent of asset management firms support the development of an anonymised, all-to-all central limit order book trading system for bonds, Tabb reports.
But almost none of the dealers are prepared to back this, Tabb’s findings show.
“Why would we provide liquidity for the whole Street?” an unnamed sell-side respondent asked Tabb. “I have a limited amount of balance sheet. Why would I care if XYZ bank wants to get out of a position?”
There’s a dangerous and increasing imbalance in the markets: a rising pile of debt held by funds promising instantaneous liquidity; and a bridge for those seeking to exit their positions that is capable of carrying less and less weight.
It’s hard to argue against the measures taken post-crisis to make banks safer. But the imbalance between bond fund inventory and dealer capacity is getting steadily worse. For the time being the cracks are being papered over by government intervention in the form of near-zero interest rates and quantitative easing, which keep the yield hunt going. Experience, though, tells us that when a market accident is waiting to happen it eventually does.