You are here

Wild markets bring equity risk premium debate to the fore

While most agree that the concept is useful, there are different views on how to derive it. BY MELISSA TAN


IN the wake of the volatility roiling Asian stock markets in recent times, equity risk premiums in the region have shot up in tandem. Or have they?

As it turns out, that depends on what you think "equity risk premium" (ERP) refers to. The issues surrounding its computation and usage can be surprisingly nebulous.

Industry practitioners, let alone academics, have yet to reach a consensus on the computation of ERP for a country's stock market and on how the figure should be used in investment decisions, according to market analysts and participants who spoke to The Business Times. The figures that have been bandied about for Asian markets, for instance, range from over 4 percentage points to nearly 8 percentage points.

As finance professor and valuation guru Aswath Damodaran lamented in a March 2015 paper: "Equity risk premiums are a central component of every risk and return model in finance ... Given their importance, it is surprising how haphazard the estimation of equity risk premiums remains in practice."

Fortunately, there is one thing that everyone seems to agree on: the concept's general meaning. The ERP for a stock market is the excess return that investors demand from holding stocks in that market, rather than a theoretically risk-free instrument such as 10-year US Treasuries. That implies that when fear rises, the equity risk premium follows suit.

But beyond this, three broad camps emerge.

Some analysts say short-term changes in stock market risk premiums can be meaningful, while others believe investors should only look at a long-term figure based on at least five decades of data - something that most stock markets in Asia happen to lack. A third group questions the very concept of ERPs, arguing that this figure perhaps shouldn't factor into asset allocation choices at all.


Analysts who track recent changes in ERP say that the figure has shot up in recent weeks. Culprits for the spike include a flight to safety among investors plus heightened fears over China's economy, the US Federal Reserve's first interest rate hike after years of cheap money and a commodities rout that has lasted longer than a year, they note. These analysts also highlight the importance of the ERP in making asset allocation decisions, saying that it can be a useful gauge of fear in the market and of the attractiveness of stock valuations.

"The ERP ... is one of our favoured measures of the degree of fear or uncertainty embedded in markets and the attractiveness of equities as an asset class," says Yiran Zhong, Asia ex-Japan equity strategist at Nomura, who computes the figure as the spread of equities' consensus forward earnings yields over the US 10-yr Treasury yield. Her ERP estimate for the MSCI Asia Pacific ex-Japan benchmark index troughed at around 520 basis points (bp) in May and has since spiked up to nearly 680 bp above the 10-yr Treasury yield of 2.1 per cent, the highest since around mid-2013.

"With this, the current level of ERP represents a 1.2 standard-deviation level of compensation for risk-taking - versus the post-2001 average ERP of 460 bp - leaving opportunity for equity upside from both earnings growth and multiple rebounds in months ahead as global growth holds up."

Delphine Georges, an analyst at fund manager Amundi, also says the ERP associated with a market is "a fundamental input in asset pricing and asset allocation decisions for investors", adding that the figure is particularly relevant for investors with medium- to long-term holding periods. She estimates that the equity risk premium went up globally by around 50 bp after the recent corrections in world markets to reach 4.50 per cent for developed markets. For the broader emerging markets, the ERP "is in attractive territory", she adds, though she does not specify this figure. To get her ERP estimate, she works backwards from current stock prices to compute the implied discount rate for dividends, then subtracts the risk-free rate. For Asia, she says the region's aggregate ERP of "around 8 per cent" is "particularly interesting" and is so high mainly because of Chinese stocks. "We believe that the equity risk premia should remain elevated in countries where firms are highly leveraged."

James Cheo, investment strategist for Asia and the Middle East at Barclays, says ERP estimates for Asian markets such as Singapore can vary wildly due to the lack of enough historical data. "One rough estimation of the ERP is by taking the difference between the forward earnings yield (or inverse of forward PE) and the US treasury yield. By that calculation, Singapore's equity risk premium was around 4 per cent before China's yuan devaluation, but since then, the risk premium has risen to around 5-6 per cent," he says.

Prof Joseph Cherian, director at the Centre for Asset Management Research and Investments at NUS Business School, says that Singapore's ERP mirrors that of the US. The US figure was around 5.8 per cent in July and went up to around 6.3 per cent in September after China's yuan devaluation in mid-August and Singapore's numbers are likely to be the same, he noted.

These numbers are based on a complex ERP model developed by NYU's Prof Damodaran, which takes into account a bunch of factors including a country's credit default swap spread and the ratio of standard deviations of its equity and bond market movements.


Other analysts believe that only long-term ERPs are valid, and that they are best used by investors with a long-term horizon.

One prominent example of such an investor is sovereign wealth fund GIC, which does indeed rely on ERPs. It said in a 2012 article that the ERP was "possibly the most important risk premium in financial markets", adding: "While investors may disagree over the exact definition of the ERP, most do not dispute that it exists." It takes patience to harvest risk premiums "because the extra returns accumulate slowly and unpredictably over time, but GIC's long-term investment approach allows us to do so", GIC said. Historically, the global ERP has been around 4-5 per cent, it said, though it did not specify how that number was obtained or how it computes ERPs.

Patrick Yau, who heads Singapore and Malaysia research at Citi Research, suggests that there's not much point in looking at short-term ERPs. "The whole concept of ERP is that it's got to be long-term ... it's a long-term indicator. By that definition anything that happens recently is not going to create much of a change in ERP. The recent volatility is too short to move the needle by much if one is thinking about 30-50 years."

However, Elroy Dimson, Paul Marsh, and Mike Staunton of London Business School suggest in a joint e-mail that there is a way to meaningfully calculate ERPs for individual markets without a long history. "It is frequently assumed that the historical premium - if calculated over a long enough period - provides a good estimate of the premium investors can expect in the future. We don't fully subscribe to this view, although we agree that history should be our starting point."

They calculate that the historical premium was 4.3 per cent averaged over the period from 1900 to 2014 but reckon the future global ERP is more likely to be about 3.5 per cent. This is after adjusting for what they believe is "non-repeatable luck" that favoured equity returns in the later half of the 20th century - stemming from things such as the end of the Cold War and rapid technological progress. Based on that 3.5 per cent baseline, they think that an individual market's ERPs can be found by adjusting for its riskiness, or beta, relative to the MSCI world equity index. "The beta of the Singapore market, for example, is around 1.2. This means that when the world index increases by 10 per cent, Singapore tends to rise by 12 per cent ... So Singapore is riskier than the world index, and its equity risk premium would probably be closer to 4.5 per cent ... Although the short-term equity premium is likely to have risen recently, the likely impact on the expected long-run equity risk premium will be quite small."


Naturally, analysts are quick to add that the ERP cannot be used on its own for investment decisions but should be interpreted within a larger matrix of indicators, and possibly taken with a pinch of salt.

Says Dr Cherian: "For the lack of anything better, and given I am a 'quant' (or have such proclivities), I prefer a scientific approach to asset allocation, valuation, investment decision making ... So I like the Damodaran approach of using ERPs for such purposes and calculations, and think it is very useful. That said, it must be used with deep care, and one must be cognisant of the 'measurement errors' in any quantitative modelling, which can lead to sub-optimal decisions. Other folks like a more qualitative approach, others prefer to lick their finger and hold it to the air, others follow a blended qualitative-and-quantitative approach."

Mr Yau also thinks that for investors today, a factor that is perhaps more significant than ERP spikes is the growing cost of capital as a whole on the back of rising interest rates. "Whether it (ERP) helps you make a better call - by itself I doubt it, but the value is that it can help create a framework that indicates if a firm is doing well or not over a longer time frame.

"It's not that ERP has changed too much, it's WACC (weighted average cost of capital). The debt side is much more interesting because debt has been very cheap over the past five years and that may have caused a lot more companies to lever up."

Other market watchers go a step further and say investors should ignore ERP.

"I am not a believer in equity risk premium," says Bordier chief investment officer Bryan Goh. "The equity premium over bonds or cash is not very stable over time and therefore it is not useful in asset allocation decisions. Equity returns are fairly stable over time, and I am being generous with that appraisal, but bond returns are not only not stable over time, they are currently artificially inflated by central bank policy. The ERP is therefore very sensitive to central bank policy since they exert disproportionate influence on bond prices.

"Even if central banks were not intervening every day in bond markets, the ERP has been unstable. The ERP is an interesting academic theory based on the efficient market hypothesis whose validity is itself questionable."

An earlier version of this article incorrectly used "his" to refer to Nomura analyst Yiran Zhong. It should have been "her" instead. The article above has been revised to reflect this.