Two recent papers provide useful techniques for “nowcasting” EM economies. The first uses “dynamic factor models” with high frequency indicators to estimate GDP growth in countries with scant and noisy data. The second uses seasonal adjustment with modifications for time-varying holidays that can track underlying trends in China and other countries with lunar year and Islamic holidays.
A two-decade empirical study shows that bond and equity market prices are more likely to “jump” on days with U.S. economic data releases. In particular, surprises in news announcements tend to lead to higher volatility and larger price moves. The impact of key data surprises on bond markets seems clearer and simpler. The impact on equity markets depends on the state of the business cycle
A common fallacy is that current account deficits measure dependence on external financing. In reality, external balances and cross border financing are only vaguely related. Vulnerability to “stops” in financial flows does not depend on trade and capital flows (“net concept”) but only on the volume and origin of financing (“gross concept”). Currency crises are not about current accounts that need to adjust, but about funding gaps that need to be closed.
When long-term government bond yields are low enough, further declines can ‘feed on themselves’. European insurance companies and pension funds are plausible catalysts. The duration gap between their liabilities and assets typically widens non-linearly when yields are low and compressed further, triggering sizeable duration extension flows.
Convenience yield can be interpreted as a leasing rate for physical commodities. Returns on convenience claims are premia earned by investment strategies for providing this leasing service. An empirical analysis suggests that they depend on risk factors related to other asset classes, however. The inertia in these risk factors seems to help predicting returns on convenience claims.
An academic paper looks at cross-country relative-value equity strategies. It concludes that [i] relative conventional factors might create alpha and [ii] relative local country equity index returns are uncorrelated with currency returns (and hence the two could be independent value creators).
An ECB working paper explains formally why senior decision makers are unprepared for crises: they can only process limited quantities of information and rationally pay attention to rare events only if losses from unpreparedness seem more than inversely proportionate to their rarity. The less probable a negative event, the higher the condoned loss. Inattention gets worse when managers bear only limited liability.
Understanding statistical learning is critical in modern markets, even for non-quants. Statistical learning works with complex datasets to forecast returns or to estimate the impact of specific events. The choice of methods is key: they range from simple regression to complex machine learning. Simplicity can deliver superior returns if it avoids “overfitting” (gearing models excessively to specific past experiences). Success must be measured in “out-of-sample” predictive power, after a model has been selected and estimated.
An HKMIR paper explains benchmark index changes and shows their significant effects on mutual fund flows and international capital flows. Importantly, there is evidence for benchmark changes leading to an outperformance of upgraded assets, both at the time of announcement and the time of actual index adjustment.
A Banca d’ Italia paper reminds us that the market impact of economic data surprises depends on the state of the economy and forecast diversity. In particular, the surprise impact tends to be greater, when predictions are tightly clustered around a ‘consensus’. Conversely, uncertainty seems to help preparing markets for shocks.
