Predicting Bond Default Rates: Martin Fridson’s Method


Martin S. Fridson, CFA

The CFA Institute Annual Conference is an unrivaled opportunity to access high-quality, unbiased educational content that equips investment professionals with the latest thinking on critical industry issues. The 70th CFA Institute Annual Conference will be held in Philadelphia on 21–24 May 2017.

Behavioral finance has cataloged an array of self-inflicted errors that can work against investors’ best interests — most often, these errors stem from functional but imperfect mental shortcuts that the brain applies in order to avoid serious cognitive effort. Successful investors understand that the market conventions arising from these kinds of shortcuts can be deeply flawed, and they recognize the importance of examining prevailing assumptions in greater detail. One such example is Martin S. Fridson, CFA: at the 69th CFA Institute Annual Conference in Montreal, he asserted that the majority of market participants are regularly projecting the high-yield default rate incorrectly.

“Most projected default rates in the marketplace, and you can read about this in the press, are based on a calculation method that takes the spread over Treasuries and subtracts a fixed illiquidity premium,” Fridson said. “This is wrong for a number of reasons.”

First, the illiquidity premium is not fixed — liquidity varies greatly depending on the issue and environment. Second, the overall spread can be the same on two dates, but with very different concentrations of default-prone issues in the index. The standard method doesn’t have internal consistency and may at times not even make sense. In 2007 for example, it said the market was expecting a negative default rate. In addition, the standard method implicitly assumes the market assesses no risk premium beyond 365 days (e.g., no default premium at all on day 366).

The Correct Methodology

The right method to determine the market’s default rate forecast assumes bonds that will ultimately default are already trading at distressed levels of over 1,000 bps over Treasuries. Fridson plotted the historical distress ratio, or the number of bonds trading at distressed levels, against the distressed default rate, or the percentage of distressed issues defaulting within 12 months, for the period 1999–2014. As the graph below demonstrates, the distressed default rate declines as the distress ratio rises.

Distressed Default Rate Declines as Distress Ratio Rises (1999–2014)

Distressed Default Rate Declines as Distress Ratio Rises (1999–2014)

Sources: BoA Merrill Lynch Global Research, used with permission; Moody’s Investor Service
Based on BoA Merrill Lynch US High Yield Index and BoA Merrill Lynch US Distressed High Yield Index

Using a regression equation, the distressed default rate percentage equals -0.3031 (a declining slope) x distress ratio +35.50 (y-intercept). This rate is then multiplied by the Distress Ratio to produce the market’s true expected default rate.

Using the most recent data as of 8 May 2016, Fridson calculated the distress ratio at 19.8% (14.0% excluding commodities), which yields a forecast for the high-yield universe of 5.8% (4.4% excluding commodities) — exactly equivalent to Moody’s forecast of 5.8%. (Moody’s uses an econometric model incorporating an economic forecast that is usually close to the consensus forecast.)

“Typically when the default rate is 1% or more above the Moody’s forecast, it is a good time to own distressed bonds,” Fridson said. “Similarly, if the bonds are priced 1% or more below Moody’s, then the distressed bonds are priced too tightly (i.e., a signal to sell).”

Rising Default Rates: Temporary Surge or Something Bigger?

At the beginning of 2016, investors worried that we were starting down a path of rising default rates and moving above the 4.5% historical average. To determine where we are in the default cycle, Fridson examined the period from 1983 to 2016 and looked for similarities.

Fridson cautioned against comparing the last cycle to today’s environment. “During the Great Recession, the default rate went from near zero to a record default rate of near 14% and back down to about 2% in just two years — a physical impossibility,” he said. “The Fed came in and reliquefied so aggressively that many companies were miraculously saved.”

A more appropriate comparison for today may be the “mini surge” of 1985–1986 when energy prices plunged. “You had a high level of defaults in the energy sector, but not the rest of the economy,” Fridson said. “In the subsequent years, default rates receded then shot up again during the recession of 1990–1991.”

Annual US Speculative Grade Default Rate, Percentage of Issuers, 1983–2016*

US Speculative Grade Default Rate Percentage of Issuers

* Projected for 2016
Yellow areas reflect recessions.
Sources: Moody’s Investors Service, National Bureau of Economic Research

Is the Current Economic Expansion Long in the Tooth?

Given the default surges typically coincide with recessions, Fridson looked at the historical length of economic expansions and the high-yield market’s current estimate of the probability of recession.

Historical evidence provides some reason to anticipate a recession in the United States by 2017. The average expansion in the last three economic cycles was 95 months, with the longest expansion running 120 months. Recently former Treasury secretary Larry Summers predicted a 50% chance of recession by 2018. If the US economy reaches December 2017 without a recession, that will represent an expansion of 102 months, exceeding the long-run average.

Fridson used the historical average option-adjusted spread (OAS) of the BofA Merrill Lynch US High Yield Index (1996 to 2015) to determine the high-yield market’s near-term assessment of the probability of recession. Over the period, the mean OAS for high-yield bonds was 1019 bps and 520 bps during non-recessionary periods.

Today, excluding commodities, Fridson noted, the probability of recession is just 9% for the next 12 months (566 bps as of 6 May 2016). Commodities industries are excluded because they are already in recession (with a current OAS greater than 1,019 bps).

“You shouldn’t use the OAS proxy as your only guide for portfolio positioning, but it is one consideration,” Fridson said. “If you think the probability of recession is greater than what the market is predicting, then you may want to underweight high yield.”

Receive updates about the conference by subscribing to the CFA Institute Annual Conference blog.

All posts are the opinion of the author. As such, they should not be construed as investment advice, nor do the opinions expressed necessarily reflect the views of CFA Institute or the author’s employer.

This post was updated on 3 March 2017 to correct the chart labeled “Annual US Speculative Grade Default Rate, Percentage of Issuers, 1983–2016”.

Photo courtesy of W. Scott Mitchell

This entry was posted in Fixed-Income Management, Investment Topics, Portfolio Management and tagged , , , . Bookmark the permalink.

2 Responses to Predicting Bond Default Rates: Martin Fridson’s Method

  1. Art says:

    Hi, in the second chart, the middle yellow bar looks misplaced? Should fall in 2001?

  2. Peter M.J. Gross says:

    Hi Art,

    You are correct, the middle yellow bar was appearing in the wrong section of the “Annual US Speculative Grade Default Rate, Percentage of Issuers, 1983–2016” chart. This was due to an unintentional transcription error, and it has now been corrected.

    Thanks for catching it!


Leave a Reply

Your email address will not be published. Required fields are marked *