For another example, please see our June 2, 2014, research brief Banks Pulling Out of Commodities … But for How Long?

The road to hell, it has been said, is paved with good intentions, and nowhere is this truer than in the byzantine world of modern financial regulation. The 848-page Wall Street Reform and Consumer Protection Act (commonly referred to as Dodd-Frank) and the oft-revised Basel III Accord have dramatically changed the investment industry, with one result that dealers have sharply reduced inventories of corporate and Agency debt. The concurrent rise in investor bond holdings through funds and ETFs begs the question of what will happen when investors look to sell.

A Short History of Financial Regulation

According to an analysis done by The Economist and law firm Davis Polk, the 11-page section of Dodd-Frank that refers to the “Volcker rule”—intended to restrict banks’ ability to engage in proprietary trading—subsequently morphed into a 298-page proposal, administered by five separate federal agencies, with 383 “explicit questions” for banks that break down into 1,420 “subquestions.” Davis Polk also created an “interactive rule map” for its clients that included an astonishing 355 different steps. As The Economist drily noted of the overall law, “there is an ever-more-apparent risk that the harm done by the massive cost and complexity of its regulations, and the effects of its internal inconsistencies, will outweigh what good may yet come from it.”

Dealer Inventory and Fund Assets
January 31, 2005 – July 31, 2014

But it is not just the law’s overwhelming complexity causing problems in credit markets. Rather, in classic fashion the noble intent of the Volcker rule—to prevent banks that enjoy an implicit government guarantee from abusing that backstop to take very risky, but potentially very lucrative positions (i.e., “heads I win, tails you lose”)—has led to unintended consequences. In short, regulations intended to address the 2008 crisis by tamping down banks’ proprietary trading, as well as requiring them to hold more capital against risk assets, have had the perverse effect of forcing banks out of their traditional role of market makers in certain areas, which will likely exacerbate problems when prices turn down.

This is not only a US problem. As the Bank of England (BOE) stated in its most recent Financial Stability Report (FSR): “Changes to banks’ business models, brought about in part by regulations to improve their resilience, have seen a number of banks withdraw fully or partially from market-making and proprietary trading activities … With banks’ reduced capacity to absorb shifts in supply and demand for securities, market prices may be more volatile in response to shocks.” (The European Central Bank [ECB] noted the same risk in its FSR as well.)

Further, the new Basel III rules on liquidity and capital have made certain low-margin (but very important) areas like repurchase agreements and Agency trading much less attractive by imposing an overall capital requirement on banks. In simple terms, while banks are required to hold only minimal capital against repo and agency trades, such assets still count toward total assets,against which banks must hold at least 5% of capital. Thus, banks are effectively incented to back away from these areas in favor of more lucrative lines such as mortgages and other loans.

Dealer, Where Art Thou?

The effect of these regulations can be seen in the precipitous decline of dealer inventories of not just corporate bonds (both investment grade and high yield), but also in Agency bonds issued by Fannie Mae and Freddie Mac, in contrast to growing Treasury inventories. Dealer inventories differ from inventory at other businesses; a decline in dealer inventory does not usually indicate pent-up demand to “restock shelves,” but more likely the opposite—that buying appetite is, and will remain, suppressed. In other words, when bond investors someday decide to sell, it is unclear who will be willing and/or able to buy. This disconnect has been exacerbated by the concurrent rise in investor bond holdings; while trading volume has risen slightly in recent years, it has lagged the explosive growth in fund assets.

US Dealer Inventories
December 31, 2000 – July 31, 2014

Trading Volume of US Investment-Grade Bonds
2005–13

In short, investors have plowed an enormous amount of money into corporate bonds through vehicles that offer daily (or in the case of ETFs, real-time) liquidity, even as the dealers that have traditionally acted as market makers—i.e., willing to stand on the other side of the trade when necessary—have been shedding such assets due to regulatory constraints. It does not take a great deal of imagination to come up with a scenario where investors look to sell … and there is no one there to buy (commonly referred to as “bids wanted” in Wall Street vernacular).

Isn’t this Old News?

Worries about the mismatch in dealer inventories and investor purchases are nothing new of course—observers have been warning about the possibility of such troubles since Dodd-Frank’s 2010 passage. In addition, some data seem to paint a more benign picture. Corporate bond funds, for example, currently hold nearly 9% in cash, which seems to offer plenty of cushion to meet redemptions should investors sell en masse—J.P. Morgan notes that even in the bond annus horribilis of 1994, quarterly redemptions never exceeded 6% of fund assets. And according to the Investment Company Institute, bond fund assets now represent 22% of assets for the total US mutual fund industry, right in line with their post-1984 average.

All true. But it is remarkably easy to poke holes in each of these comforting storylines. The fact that warnings about declining debt liquidity have thus far failed to pan out is not evidence such warnings were wrong; indeed, both the 2008 and 2000 crashes were preceded by several years of “fallacious” warnings. More broadly speaking, all such analyses are based on the assumption that the future will be similar to, and certainly no worse than, the past. It’s easy to imagine an analysis done in 1993 that showed corporate bond funds had never suffered any major redemptions (much as several firms, in addition to former Federal Reserve Chairman Ben Bernanke, pointed out just prior to the 2008 crisis that US home prices had never fallen on a nationwide basis). Likewise, the fact that bond fund assets make up a similar share of total fund assets to their historical average is only meaningful so long as one ignores equities having more than tripled off their 2009 lows. Bond fund flows have until recently been strongly positive—retail investors have poured more than $1 trillion into US bond funds since early 2009—and the recent reversal in high-yield flows, which occurred as soon as performance turned down, shows how fickle such flows can be.

The Fed’s recent trial balloon about imposing exit fees on corporate bond funds is either indicative of heightened worries about an impending rush out of bonds, or a remarkably tone-deaf attempt to start a conversation about the issue. Or perhaps both. Either way, the Fed’s decision to raise the matter, combined with bond funds having grown ever larger and yields have compressed even further, suggests investors should be watching this space.

Where Could We Be Wrong?

A collection of loose tinder does not a fire make. So long as rates stay low and markets quiescent, bond funds can continue to collect assets even as dealers shed inventory. Further, governments not only created this issue, but also have the power to “fix” it. Not, to be clear, in the sense of more clearheaded regulations—for which we hold out little hope—but rather in the likelihood that they continue their pattern of frequent and sustained market interventions.

To wit, would anyone really be surprised if the Fed stepped in to buy corporate bonds—or gave banks money to do so—in a market panic? If there is one thing we have learned over the past few years, it is that government and central bank officials (but we repeat ourselves) are willing to do “whatever it takes” to support prices of risk assets.

Another, more benign possibility is that markets will adapt to the new rules. For example, one area that has seen some improvement is trade size. While block trades (>$5 million) still make up a smaller percentage of the market than they did a few years ago, they seem to have stabilized, and the average trade size for the overall market has also rebounded.

Trade Sizes of US Investment-Grade Bonds

The Bottom Line

In classic fashion, regulations intended to address excesses from the most recent financial crisis have laid the groundwork for a different problem next time round. The retreat of dealers from market-making in various debt markets is clearly a concern if and when prices turn down, particularly given the headlong rush by retail investors into bond funds with daily liquidity and, even more worryingly, ETFs. While this is not a new concern, the Fed’s decision to raise the issue of exit fees—as well as the ECB and BOE’s comments in their respective FSRs—seems a potential red flag.

Contributors

Eric Winig, Managing Director
Ramin Sobhany, Investment Associate

Exhibit Notes

Dealer Inventory and Fund Assets
Sources: EPFR Global, Federal Reserve, and Thomson Reuters Datastream.
US Dealer Inventories
Sources: Federal Reserve and Thomson Reuters Datastream.
Notes: Data are monthly. Data for corporate bond inventories start July 31, 2001.
Trading Volume of US Investment-Grade Bonds
Sources: Barclays, Bloomberg L.P., and FINRA TRACE.
Order Sizes of US Investment-Grade Bonds
Sources: MarketAxess and FINRA TRACE.
Notes: Micro < $100,000; Odd = $100,000 – $1,000,000; Round = $1,000,000 – $5,000,000; Block > $5,000,000. Volume percentages are monthly through July 31, 2014. Average trade size data are annual, with data for 2014 through July 31.