Why has investing in Taiwan been such a loser trade over the decade?

Under-promise and over-deliver is a standard recommendation of the self-improvement literature. The Taiwan stock market, though, is taking the opposite route. Over the past decade, the TAIEX rose 45%, compared with an average gain of 120% in regional stock markets and an increase of 240% in the Korean market. This note explores the reasons for Taiwan's poor stock record and asks whether it is likely to persist.

Can subpar economic growth explain poor stock performance? Far from it. Over 2001-2010 Taiwan's economy grew at an annual rate of 4.5%, or well ahead of Korea's 4.2%. This is a peculiar finding, given that growth projections are a key ingredient into portfolio managers' decisions on asset allocation-be it across asset classes or across countries.

Could it be that the choice of base year affects the results? Taiwan was hit harder by the 2001 burst of the dotcom bubble than Korea, given its greater reliance on IT exports. Hence, its superior growth performance post-2001 could reflect a stronger bounce back, rather than stronger fundamentals. To address this question, I compare the levels of GDP (instead of the growth rates) by converting them at purchasing power exchange rates.

Remember, this is the exchange rate that makes a Big Mac cost the same in Taipei and Seoul and can be thought of as a real exchange rate. I also use GDP per capita (instead of GDP) to make sure demographics are not behind the difference in growth performance. As shown in Figure 3, Taiwan's economy continues to outperform Korea's, despite a higher base at the outset of the observation period.

Could it be that Taiwanese workers appropriated a larger chunk of GDP than their Korean counterparts, leaving a smaller share for stock owners? I have written extensively on the topic and, if anything, the opposite is the case. Taiwan is in fiercer competition with China than Korea, given closer trade ties, greater cultural and linguistic proximity, and a product mix that is easier to emulate by mainland China. As a result, Taiwanese workers needed to exert more wage restraint to keep their jobs.

Real wages in Taiwan remained flat over the past ten years. In Korea, real wages came down markedly in the wake of the global recession, but are still 8% above 2001-levels. National Accounts data tell the same story. In Taiwan, the share of GDP appropriated by wage earners fell from 48% in 2001 to 46% in 2010. In Korea, labour's share in GDP rose from 43% to 46.5% during this period. If Taiwan's GDP outperformed Korea's, but both labour and capital incomes failed to keep pace with Korea, where did all the money go?

Could it be that stock earnings didn't do so badly over the past decade, but that most of the performance was already priced into Taiwanese stocks at the outset of the decade? Put differently, could it be that the Taiwanese stock market was overvalued in 2001 relative to its Korean counterpart? To answer this question, I compare the cyclically adjusted price-earnings ratio of the two markets (defined as the stock price divided by the 10-year trailing average of earnings per share). Taiwan's 2001 stock valuation, at 30 times historical earnings, was indeed much more demanding than Korea's, at 22 times historical earnings.

To quantify the effect that initial overvaluation had on successive stock returns, I break down changes in the stock index into changes in the (simple) price-earnings ratio and changes in the earnings per share. I then calculate how much of the TAIEX underperformance (relative to the KOSPI) can be explained by the underperformance of the price earnings ratio (initial overvaluation) and how much by the underperformance of earnings per share (fundamentals). It turns out that about half of the TAIEX underperformance over 2001-2010 can be explained by initial valuations.

The other half of TAIEX underperformance is due to weaker earnings per share. How can this be reconciled with Taiwan's better GDP performance and greater wage restraint? One possibility is share dilution. There are four sources of GDP growth: equity-financed capital investment, debt-financed capital investment, increased labour input, and productivity growth.

Among those four, only the last three translate into higher earnings per share. In contrast, raising more equity for capital investment will boost GDP and aggregate earnings, but not earnings per share. However, the disconnect between Taiwan's GDP and stock performance is not due to share dilution. Taiwan's earnings still underperform-in stark contrast to its GDP.

This leaves us pretty much with the last possibility that listed companies are not a representative sample of the Taiwanese economy. At first sight this hypothesis does not seem to hold water. Taiwan sports the largest stock market in Asia (excluding city states), when measured against the size of its economy. But, listed companies are indeed a poor representation of the Taiwanese economy, as many have large production operations in mainland China. This is reflected in Taiwan's stock of outward FDI which amounts to 50% of GDP, compared with 15% in the case of Korea.

Given the extent of outsourcing, mainland wages are more important for Taiwan's stock performance than Taiwanese wages. And, Chinese wages have tripled over the past 10 years. In general we should be able to monitor both onshore and offshore activity through Taiwan's GNP, which adds to GDP the factor incomes earned abroad. Unfortunately, GNP data is not reliable and so there is no way to formally reconcile micro and macro data.

However, there is other evidence which suggests that overseas operations are indeed to blame for Taiwan's poor stock performance. The earnings of various sectors over 2001-2010 are taken from the top 5 listed companies in each sector. Three sectors are purely Taiwan-based, namely construction, steel, and semiconductors (the latter were forbidden to produce in the mainland until recently for strategic reasons).

Those Taiwan-based industries were much more profitable than assembly industries with major mainland operations. The main exception is the "other electronics" sector which includes the likes of Hon Hai Precision Industry and Foxconn Technology. However, this sector is riding a consumer product cycle which allows it to offset shrinking margins with larger volumes.

To sum up, Taiwan's stock market underperformed regional peers over the past 10 years despite stronger GDP performance. Half of this underperformance was due to ambitious valuations at the outset of the observation period. The remainder was due to rising wages in mainland China which cut into the profits of offshore affiliates (which are not covered by GDP data).

What do these findings imply for future stock market performance? Mainland wage pressures are unlikely to go away. From a cyclical perspective, monetary and fiscal policies are still loose. But, also from a structural perspective wage pressures are likely to persist, as productivity rises, living standards improve, and goods move up the value-added chain. Add to this Beijing's goal of strengthening wage growth and private consumption, and you can see significant headwinds facing Taiwanese companies.

How to steer these headwinds? Invest in sectors that are up the value-added chain or have a Taiwanese production base. Examples of the former are cloud computing or biotechnology, examples of the latter are tourism and retail both of which are poised to profit from rising mainland tourism. Given stock performance over the past decades there is scope for positive surprises.