In Valuation, we often come across this term “risk-free rate” also known by its nickname - “Rf”. In general usage, most valuers use the 10-year G-Sec Yield published by FBIL as the risk-free rate. Which makes sense. The risk-free rate is supposed to be least risky investment one can find, and we add a risk premium on top of that for valuation. Right? No.
There’s a difference between a “risk-free” rate and the “least risky” rate. While Government Securities (G-Secs) may be the least risky rupee denominated investments in the world, they are certainly not “risk-free”. After all, we’re talking of Rf and not RL. In fact, on 1 Apr 2020, the sovereign rating of India assessed by Moody’s was Baa2. The typical default spread for this rating was 2.82%. Further, the average credit default swap (CDS) rate for dollar denominated Indian bonds was 2.78% higher than the CDS of Swiss bonds.[1] Hence we see that the Indian G-Secs are far from risk free. There are two possible ways to adjust for the difference.
- We can take the 10-year G-Sec rate and reduce the sovereign default spread to arrive at the risk-free rate. For example, on 1 Apr 2020, the default spread for India was 2.82% against a 10 year G-Sec yield of 6.73%[2]. Hence, the risk-free rate of return for India was 6.73% - 2.82% = 3.91%
- We may start with a 10-year Government bond yield rate for a riskless economy, and add the CDS of the Indian economy to that. The inherent assumption here is that there exists a riskless economy, which is not true. However, we would rather base our judgement on an incorrect assumption rather than do away with the measurement altogether. Let us assume Switzerland to be a riskless economy with very little chance of sovereign default. The CDS spread of India over Switzerland comes to be 2.78%. Given that the yield of Swiss Government Bonds on 1 Apr 2020 was -0.37%[3], the risk free rate for India should be -0.37% + 2.78% = 2.41%
The two calculations give us very different numbers. However, we see that the “risk-free” rate is very different from the 10 year G-Sec yield which is not inherently riskless. Still it is common practice to assume the 10 year G-Sec yield. In fact, why use the 10 year G-Sec bond yield at all? Aren’t 90 day T-bills riskless by the same sovereign guarantee? In fact, since risk lies in an uncertain future, lower duration T-bills should be less risky and hence more “risk-free” than 10-year G-Secs. So why do we use the 10-year G-Sec yields?
Purist valuation practices suggest that the duration discount rates should be equal to the duration of the cash-flows. In the perfect sense of the definition, each cash-flow should have a unique risk-free rate equal to the duration of the cash-flows. In going concerns, the bulk of the value comes from the future terminal value of cash-flows. Hence, it is logical to assume that the longest duration G-Secs used for the discounting of all the cash-flows will result is very little error in calculations. However, we must not forget that in cases of SPVs with limited time-frame of operation, such as many real estate or infrastructure projects, the 10 year G-Sec rate may not be ideal for calculation of the risk-free rate.
An argument in favour of using the 10 year G-Sec rates is the Rf on its own is meaningless. What we intend to find is the cost of equity (Ke) which also incorporates a market risk premium. As long as we adjust the market risk premium correctly for our risk-free rate we calculated, our Ke will come out correctly. But that’s a different discussion.
For now we may change the way we use Rf, or we may stick to the conventional practice of using 10 year G-Sec rates as Rf, but we should remember the baggage that comes with it. The 10-year G-Sec yield is not risk-free, but is a good starting point to find the correct risk-free rate.
No comments:
Post a Comment