Stock indexes of emerging markets (EM) have been performing poorly against developed markets (DM) over the last few years. Interestingly enough, the underperformance was not explained by traditional drivers:
i. Currencies (I use Asia here through ADXY) remained broadly stable against the USD while past experience would have suggested weaker currencies (chart below, left) or a stronger EM relative performance.
ii. Above all, the rise in global trade, albeit having decelerated sharply throughout 2013, should have justified higher relative valuations for EM stocks (chart below, right).
iii. VIX and EM/DM: the link between relative appetite for EM stocks and risk aversion has vanished since late 2011. As can be seen below (left chart), the relationship seems to remain locally verified. A focus on the recent past shows yet that, since mid-2012, the relative return of EM to DM stocks is no longer linked to VIX. We may discuss the validity of VIX as a good gauge of risk aversion, but in any case the outperformance of DM stocks cannot be explained.
In addition, the widening gap in PER levels did not lead to a mean reversion of the EM/DM ratio. One simple explanation could be that the widening in Price Earning ratios were not accompanied by a widening of the ROE of the same amplitude. Interestingly enough, as can be seen below, the ROE (Return on Equity) of EM listed companies fell during the period and remains only slightly above that of DM markets. This fall in absolute and relative ROEs would mean that return on capital is no longer much higher in EM countries than in DM economies: further proof of inefficient allocation of resources and a necessary adjustment of many countries' growth models (this would also call for more discrimination among EM countries).
This would be enough to justify the underperformance of EM markets. I explore some potential explanations below:
Tapering should not be blamed, as the long-lasting under-performance started in the midst of very aggressive quantitative easing rounds. The tapering fear wreaked havoc on the link between the gold/copper ratio (it used to be a good proxy of fear/growth perception of investors) and the S&P 500 vs. EM stock relative return. I cannot say to what extent the tapering fear translated into a bubble burst for gold prices or triggered EM fears. Probably both. In any case, the metric is no longer reliable.
Monetary policy and US Yields
Historically the relative performance of EM markets relative to DM indexes was negatively related to T-Note prices (positively to long term yields): higher T-Note prices would reflect episodes of high risk aversion (safe haven flows) and/or weak US growth, which would mean weak global perspectives. This would generally translate into an outperformance of advanced stock indexes.
Conversely, in periods of smooth US growth, higher interest rates would not be synonymous of lower global liquidity, as is often thought, but rather of better global growth prospects, hence the outperformance of EM stocks. This is what the chart below highlights.
Turning to the post crisis periods, US Treasuries have performed positively and DM markets have outperformed EM stocks. Therefore, no significant regime switch is noticeable. This would suggest that the growth/higher yield explanation for EM relative performance is more convincing than the liquidity/low rate/portfolio inflows in EM story. Two major exceptions stand out in the recent past:
i. The post QE1 EM outperformance was clearly driven by the first implementation of quantitative easing by the Fed. The associated outflows from the US could explain the EM outperformance. However, the following QE would not trigger a similar pattern.
ii. The fear-of-tapering sell off of mid-2013 was also based on liquidity-drying fears. It led to an EM selloff that had, as we stressed several times, its sources in bad EM macro management policies.
As can be seen above (circled areas), disconnects are noticeable in periods of reversal (or innovation) in US monetary policy. But, they never last for long. Therefore, though this may seem paradoxical, US monetary policy may explain the recent pattern of EM/DM return differential not because of the extra liquidity being poured into emerging countries (in that case the relationship would be inverted) but because low rates are historically synonymous with DM outperformance.
As the Fed launched a soft tapering in December, higher US Treasury yields ahead could benefit and not harm EM stocks if the old relationship holds.
Local Sovereign bonds as "new" safe haven
Another explanation of the underperformance of emerging stocks during the post-2008 period would be the switch from EM stocks to EM debt (in local currency notably) - a move motivated by the end of the "original sin" (inability of many developing countries to issue debt in their own currency) and the scarcity of safe haven bonds following the Eurozone crisis.
Before I focus directly on local corporate bonds, the chart below highlights the relationship between EM/DM stocks returns and the return differential between the EMBi (EM debt denominated in foreign currency) and the GBI (total return index of local debt denominated in domestic currencies).
As can be seen below, the outperformance of DM on EM stock markets came along with that of local bonds against foreign currency denominated EM debt (adjusted for FX fluctuation).
This would be no surprise for investors who have considered local currency bonds as safer, would-be candidates for the status of safe haven. Local currency bonds issuers are those who have gained enough credibility in their macro-management so that they can issue debt in their own currency. As a result, the aggregate credit quality (rating) of the GBI stands much higher than that of the EMBi and both behave differently in periods of heightened risk aversion (this is without taking into account liquidity differentials between both markets). From this analysis I would be very reluctant to draw any conclusion on EM/DM performance from the relative behavior of dollar versus local currency sovereign bond behaviors. The only conclusion is that both safer assets (DM stocks and local sovereign debt) have outperformed during the period.
Local Corporate Debt
The most important pattern change observed over the last five years has been the rapid growth of corporate issuance in emerging countries. As a result it might be interesting to gauge whether the CBI-EM spread with the CEMBI (local currency denominated debt issued by EM corporates) tracks the EM/DM or not.
As can be seen below, even if the CEMBI/GBI return spread has tracked that of EM/DM stocks, emerging local currency corporate debt has almost always outperformed local government debt. If the EM/DM return differential widens when risk appetite is high then an outperformance of local currency corporate bonds versus local currency sovereign bonds seems natural during those times (and conversely).
Unfortunately, those co-movements cannot provide a strong explanation of the long drift downward of EM/DM relative performance.
Bottom Line: The long lasting decline the relative performance of EM stocks against DM indexes cannot be explained by the traditional drivers: VIX, global trade and emerging currencies behavior. Even relative price earnings ratios have not triggered any mean reversion in the return spread. The worsening of the ROEs of EM listed companies is probably the best suspect here, with a worsening in the marginal productivity of capital.
Interestingly, the new patterns of emerging investing (local currency denominated debt) cannot explain the relative returns of EM against DM stocks. I just identified some solid patterns (DM out performance against EM stocks comes with an outperformance of GBI against EMBI and CEMBI against GBI) but no clear cut explanation.
Lastly, and contrary to what is often believed, if the Federal Reserve manages to control the hike in long-term yields in 2014, it might not be as detrimental for EM stocks.