Monday, August 03, 2009

Deciphering Hot & Hotter Emerging Markets



Trying to further decipher whether the flow of funds into emerging markets' equity is broad based or selective, is an important question. If it is broad based, then it is primarily a big picture capital flows trend. Generally, when we get the big picture correctly, the smarter money would further drill down to specific markets. Not all markets are created equal. At any point in time, some markets will be more attractive than others based on the prevailing interest rates, growth rates assumptions, equity valuations and other lesser investing factors.

I have managed to come across a great chart by the highly respected Bank Credit Analyst, which basically explains which markets would "see the most action" in this current rally. Anyone could plot a valuation chart based on the country's prevailing interest rates and match that with the country stocks' forward or trailing PER. That would be quite one dimensional. The BCA's chrat is a lot more persuasive in that it is based on forward and de-trended (I don't even want to attempt to know what that word meant, but it sounded so sophisticated) trailing PERs, price to book ratio, and dividend yields shown relative to the average of 18 countries".

Hence it is a peer-to-peer analysis. There is one major shortcoming in that tabulation, in that it does not take into account the historical average valuation of each specific market say over the past 5 years. The BCA chart only looks at how each country is faring in valuation terms relative to one another. For example, if Malaysia's historical PER forward valuation was 18x, and HK's historical PER forward valuation figure is 15x - naturally if we just look at PER, HK will always look cheap relative to Malaysia. However if the current forward PER for Malaysia and HK are 14x and 13x respectively - HK would still be cheaper on a straight out valuation but in actual fact, we should look at its PER now compared to the historical valuation for a more nuanced and value-add commentary, whereby Malaysia's market would be the 'cheaper' one.

Besides that caveat the BCA chart looks very good indeed.

From the BCA chart, the most attractive markets NOW:
1) Czech and Thailand - these are "false favourtites" in a horse race. On valuation matric, they are the most attractive and cheapest, but their valuations have been skewed because of extreme financial distress (in Czech Rep) and heavy political uncertainty in Thailand. Hence we need to take these two out or at least regard them with a lot of conservatism.

2) HK - Its looking really good, tus explaining the surge in hot money there. Shanghai and Shenzhen are not so open to foreign funds, hence much of the hot money has been diverted to HK as proxy.

3) Poland - Cheap on valuations.

4) Malaysia - You wouldn't get many houses recommending Malaysia, but the BCA thesis puts Malaysia as highly attractive. Bank Negara's recent decision to leave rates unchanged makes things hotter still till the next BN meeting.

5) South Africa & Hungary - These are attractive but have very high rates, which indicates that money might be flowing out of these two countries. In an effort to retain capital flows, these rates are kept high. In instances like these, foreign funds regard these as tricky markets as your equity gains could be erased by a weak local currency in the end.

China is very hot even though its not the most attractive on valuations, it is a relatively closed market and the surge in bank lending (i.e. liquidity) over the past 6 months has ensured a most vibrant market.

Singapore unfortunately may be trailing the rest of the emerging markets as valuations wise, it is not that attractive. Still, trailing it may be, it will still enjoy some partying albeit much less enthusiasm.

1 comment:

hishamh said...

I'm not sure what it means in this context, but detrending is fairly simple to explain - it's simply taking as your values the fluctuations around a trend.

If you fit a regression line to a time series, and minus the actual observations against the corresponding values of the regression line, you'll get a detrended series (aka residual series). That's one way - the simplest would be to just take an average.

I don't know in this instance if he is referring to a time series of PER, or panel data (across countries).