Life mimics nature — it is cyclical. The amplitudes and lengths of the cycles vary, but human behavior remains resolutely cyclical, despite the evident errors that past cycles produced. One might think that people would modify their behavior enough to avoid past mistakes, but it is apparently built into human nature to repeat them. The particulars are never the same, but the underlying patterns are the same.
So it is in financial markets. Credit extension practices are remarkably cyclical, despite the experience of prior cycles that should discourage errors. There is even a hint of this in the word itself. Credit is derived from the Latin word credere. It has multiple translations, but they include believe, trust, and imagine. The power of belief in the financial markets is enormous, and produces both undervaluation at times and, at others, excessive valuations.
The salient example of the cycle of ten years ago was, of course, single family mortgage debt. Price appreciation of the underlying homes, financial innovation that grew like kudzu, and a modest contribution from government incentives to lend caused the underwriting errors that eventually produced the financial crisis of 2008-09. Prior to that, the cycle of 1998-2002 was centered on corporate issuers, in the form of massive overvaluation of technology stocks and serious credit errors in corporate bonds, particularly high yield.
It seems that the duration of collective memory is good for one cycle, but not two. This time the mistakes are in corporate bonds, and the markets are beginning to hint at their appearance. Certainly the level of credit creation in the corporate sector has reached levels at which the oxygen began to thin out in the past. In Chart 1 below we show corporate debt to GDP:
Chart 1 Source: Federal Reserve, Bloomberg
This would seem to argue for circumspection, but that is apparently not the conclusion of the financial markets at the moment. The perceived riskiness of a bond is its credit spread—the excess of its yield over that of Treasuries, the default risk-free alternative. In Chart 2 we overlay that series on the first one:
Chart 2 Source: Federal Reserve, Bloomberg
Credit spreads remain at the bottom end of their historical distribution, despite the elevated level of credit extension to the corporate sector. The belief underlying such valuations is apparently that such a level has not produced the same mistakes that occurred in the past. It is certainly true that default rates remain low, as one would expect with an economy performing well — which it should be, with added nourishment from tax cuts. But we also know that default rates are a lagging indicator, since businesses generally only default when they have little alternative.
It may be useful, then, to look for any indications of vulnerability. We do find a handful. That is all at the moment, but they are suggestive of some of the mistakes that may have been concealed from view by the salubrious conditions that have prevailed. Those conditions were at least in part an intended effect of central banks’ behavior (likely including spillover from the European Central Bank’s corporate bond buying program, an alternative denied the Fed by statute).
The first is Toys “R” Us. This was at one time a favorite of the capital markets as it benefited from unit expansion and growth in same-store sales. But in retail, even more than in other sectors, change is the way that the iron fist of time levels the playing field. In 2005 the company was the object of a leveraged buyout, in which debt is used to buy out the public shareholders. The new owners are private equity firms, who intend to profit after an ownership period from appreciation in the business value, with that profit fattened by the use of leverage. In this case the leverage was excessive. The debt load, together with the shift to ecommerce across retail, was the beginning of the end.
Toys “R” Us filed for bankruptcy in September of 2017. The prices of the company’s securities were remarkably resilient, however. Chart 3 shows the performance of the most widely traded bonds from the time they were issued in 2016:
Chart 3 Source: Bloomberg
Despite imminent, and then confirmed, bankruptcy, the bonds traded at prices that would prove to be optimistic. Retailers generally do not fare as well in financial distress as businesses that own durable assets other than inventory, which is often subject to markdowns. The redoubtable performance of Toys bonds is best explained by a general desperation for yield across financial markets, and a resulting hope that they could make the interest payments, rather than care in security selection. They are now trading at materially reduced levels.
The second case is General Electric, another erstwhile icon. In this case, the mistakes were expansion, particularly in financial services, management miscues and, again, a liberal use of debt. GE has spent considerable energy disposing of businesses that were accumulated during the growth phase, but a debt load remains. A series of disclosures about financial problems that had accumulated over time have caused the market to question the company’s current BBB credit rating. Even GE’s bonds reflect a revision of the markets’ view (Chart 4):
Chart 4 Source: Bloomberg
These are specific observations and may not be indicative of broader issues to come. But they both suggest that recent years have been a period of confidence that proved to be excessive, and the price revisions under way are a measure of reality in these cases. We shall see if this implies broader revisions to valuations going forward, but investors’ yield starvation in recent years probably induced them to make mistakes. At this point investors should err on the side of excellent credit quality, and even scrutinize firms that are — for now — generally perceived to be safe. It is worth recalling that, not so very long ago, GE was rated AAA.
John R. Gilbert is Co-Director of Research at Bradley, Foster & Sargent Inc., of Hartford, Connecticut.
