Common Fallacy: Misuse of Relative Valuation

 

Lately, lots of brokerage and equity research firms have been comparing price-to-earnings ratio (PE) over time to show how cheap Pakistani stocks are. While Pakistani stocks may be cheap as per historical standards, the usage of PE to justify it is completely flawed.

Fundamentally speaking, PE is a relative valuation metric; that means, it only tells you how cheap a company is relative to the sector. Even that comes with a caveat: provided that the comparison is to companies with similar characteristics – a condition which we will return to later.

PE tells you nothing about how cheap one market is to another (country-to-country comparison) nor how cheap a market is over time. To comprehend this, one has to evaluate the theoretical basis of PE.

The Gordon growth model can be used to determine value of equity: P = D1/(r - g). Dividing this by earnings would give you PE: PE = Dividend payout ratio/(r - g). Now, you have a theoretical basis to understand all the factors which influence PE.

In the formula for PE, you’d see that r and g (in denominator) are the key determinants of PE. If growth rate (g) goes up, PE becomes higher and if cost of equity (r) goes up, PE becomes lower.

Cost of equity is generally calculated by adding an equity risk premium (ERP) to the risk-free rate (base interest rates). So, if the State Bank hikes interest rates, PE ratio for PSX should automatically go down. Note that interest rates are at record-high levels, hence it makes sense that PE for KSE100 remains low. Therefore, using PE to say KSE100 is cheap right now is useless. It ignores the fundamental relationship between interest rates and equities.

Similarly, PE cannot be utilized to compare markets to one another. For instance, Pakistan’s ERP is likely to be far higher than USA’s due to various issues such as poor corporate governance or high political instability. Consequently, stable period PE would be far lower for Pakistan relative to the USA.  

So, what can PE really be used for? It can be used to make relative valuation plays. PE only tells you if a company is trading cheaply relative to its peers (sector averages) which share similar characteristics. What are those similar characteristics? They could include growth rate, dividend payout/yield and risk. This is coming directly from the PE formula used above.

Since growth rate is a significant determinant of PE, investors often opt to use PEG ratio which is simply PE divided by the growth rate of the company. On the other hand, companies with higher dividend yield or payout tend to trade at higher PE multiples as larger payouts are associated with lower risk – a tenet since the dawn of the joint-stock company.

So, dividend yield/payout ratio can be used as a gauge of risk and many investors incorporate this to a make a new metric out of the PEG ratio. However, others opt to use statistical standard deviation as a proxy of risk which is fine too.

By performing such relative valuation, investors can exploit temporary mispricing of securities in the market. This blog has previously done such analysis for HBL which yielded favorable results.

But note, if the sector itself sees a selloff due to any reason, then the relatively mispriced security will trade downwards as well. So, for relative mispricing to be corrected by the market, there has to be a significant positive trigger which was higher earnings in the case of HBL.

Now that the fallacies regarding the usage of relative valuation metrics have been clarified, how can an investor determine if the market is cheap? There is an excellent metric for this: 5-year annualized return for KSE100.

Much like the economy, markets are cyclical. A boom is often followed by a slump and vice versa ad infinitum. The above graph shows that cyclically, the KSE100 index is in its slump phase. It’s anyone’s guess when the market could return to the boom phase. But this is still an excellent indicator that KSE100 is cheap given its position in the market cycle.