Distress Risk basket enhanced in Conservative Credits

Distress Risk basket enhanced in Conservative Credits

24-04-2014 | Research
  • Jeroen  van Zundert
    van Zundert
    Researcher Quantitative Credits
  • Patrick  Houweling
    Portfolio Manager
  • Georgi Kyosev
  • Low-risk credits offer superior Sharpe ratio 
  • New risk variables added to Distress Risk basket 
  • Higher risk not rewarded with higher return

Conservative Credits

Research by Robeco and academic researchers shows that a low-risk anomaly exists in credit markets: low-risk credit portfolios earn higher risk-adjusted returns than high-risk portfolios over a full market cycle. Robeco Conservative Credits, launched in 2012, exploits this anomaly by investing in low-risk bonds of low-risk issuers. Both the research and the live track-record show that low-risk credits realize a superior Sharpe ratio compared to the credit market index.

Blijf op de hoogte van kwantitatieve inzichten
Blijf op de hoogte van kwantitatieve inzichten

Distress Risk basket

Conservative Credits uses a quantitative model to rank issuers in the low-risk bond universe. Our Distress Risk basket is the dominant driver of the model3 and measures company risk using balance sheet and market data. The main variable in the basket is Distance-to-Default, which is constructed in the spirit of the Merton (1974) structural credit risk model. Distance-to-Default measures a firm’s default risk using its equity market capitalization, equity volatility, and balance sheet leverage. Robeco research shows that a portfolio of low-risk companies, as identified by our Distress Risk basket, generates a similar return as the market, but with much lower volatility, resulting in a superior Sharpe ratio 4. Moreover, such portfolios generate higher Sharpe ratios than portfolios based on generic risk measures, such as credit rating.

Enhancement for Corporates

After an extensive research project, investigating possible additions to the Distress Risk Basket, we decided to add cash flow leverage to the basket. Cash flow leverage is measured as a company’s current debt level divided by the generated cash flow over the past year. It may be interpreted as the number of years a company needs to redeem all its debt. Obviously, we prefer companies with lower leverage over companies with higher leverage. We conducted a back-test by creating quintile portfolios: the Q1 portfolio contains the 20% firms with the lowest leverage, followed by the next 20% in Q2, etcetera, until the most levered 20% in Q5. Figure 1 shows that companies with higher cash flow leverage clearly had higher volatilities. However, portfolios Q1 to Q4 earned similar returns, irrespective of their risk, while the highest-risk portfolio (Q5) even had the lowest return. This is the right behavior for the Distress Risk basket.

Figure 1: risk-return plot cash flow leverage
Source: Robeco Quantitative Research, Barclays, Factsheet, USIG Jun93-Dec13

Enhancement for Financials

The cash flow leverage variable is only applicable to corporates and not to financials due to their specific accounting rules. Therefore, we also analysed various risk variables that are specific to financials. After extensive investigations, we selected the Tier-1 ratio for addition to the Distress Risk basket. The Tier-1 ratio is defined as the ratio of the core equity capital to the risk-weighted assets. We prefer higher Tier-1 ratios, since this means that more capital is available to cover losses on the loan portfolio of the bank. Figure 2 shows the risk and return of tertile portfolios constructed by ranking financials on this variable 5. Firms with the lowest Tier-1 ratio clearly had the highest risk, which was not compensated with a higher return. Once again, these results show a low-risk anomaly.

Figure 2: risk-return plot Tier-1 ratio
Source: Robeco Quantitative Research, Barclays, Factsheet, USIG Jun93-Dec13

More breadth

The quantitative results of our research project show that the new variables have the desired low-risk characteristics and that they have added value to the model. Additionally, we are happy with the increased breadth of the model and the added diversity of the information contained in the Distress Risk basket. This increases the quality and robustness of the low-risk screening in our model.


Due to this enhancement of the Distress Risk basket, we are confident that the Conservative Credits strategy will continue to generate attractive Sharpe ratios for our clients.


  1. Robeco research: Derwall, Huij & de Zwart, 2009, “The Short-Term Corporate Bond Anomaly”; Houweling et al, 2012, “The Low-Risk Anomaly in Credits”; Houweling, van Vliet & Wang, 2014, “The Low-Risk Effect in Credit Markets”. Academic research: Ilmanen et al., 2004, “Which Risks Have Been Best Rewarded?”, The Journal of Portfolio Management; Kozhemiakin, 2007, “The Risk Premium of Corporate Bonds”, The Journal of Portfolio Management; Frazzini & Pedersen, 2014, “Betting Against Beta”, Journal of Financial Economics; Carvalho et al., 2014, “Low-risk Anomalies in Global Fixed Income”, The Journal of Fixed Income; Aussenegg et al., 2014, “Common Factors in the Performance of European Corporate Bonds”, European Financial Management. 
  2. Whenever we mention return in this memo, we mean the excess return of a corporate bond vs. duration-matched government bonds. This is the credit-risky component of the total return of a corporate bond. The interest rate component is irrelevant for our analyses, because investors can steer their interest rate duration separately and cheaply using bond futures or interest rate swaps. 
  3. Value and Momentum are supportive factors in the model to further enhance the Sharpe ratio. The Value basket was also improved recently; this will be described in a separate note. Enhancements to the Momentum basket were already documented in our white paper “Smart Credit Investing: Residual Equity Momentum” from August 2013. 
  4. This research is document in our white paper “The Low-Risk Anomaly in Credit Markets” from April 2012. 
  5. Here we use 3 tertile portfolios instead of 5 quintile portfolio, because of the smaller number of financials compared to non-financials.