VoxEU Column Global crisis International Finance

Fickle inter-asset linkages: The need for effective policymaking

If the base rate rises, all things being equal, the exchange rate is expected to rise and bond prices to fall. This column argues that, during a financial crisis, such relationships between asset classes go haywire. When this happens, it says governments (including central banks) must provide strong signals to the market and make sure that they pick up the right signals from the market themselves.

Investors’ decisions are largely influenced by the interplay of factors such as the degree of risk-taking ability, the prevailing economic situation, market sentiment, relative and perceived scarcity of asset classes in the economy, the prevailing policy environment, and the business cycle – all of which form the basis of inter-asset linkages at some point. Yet the way these factors influence investor decisions during a normal or boom period is different from that during a crisis. There could be two reasons for this. First, during a crisis, policymaking concentrates more on dousing the fire than on maintaining inter-asset linkages – something that calls for a coordinated policy action. Second, the opacity created during a “bust” leaves market participants “blindfolded” such that even a small piece of positive information drives them to chase returns with little diligence over risks, which typically become opaque in such times. The result is that some investors keep away from markets (Wilson 2010) while many of those who participate end up chasing only “returns” and the market largely misses out on the arbitraging (the linkages) opportunity.

Crucial questions for the “asset diversification” theory

During a crisis inter-asset-class relationships become fickle and demonstrate what is known as “correlation breakdown” in simple statistical terms. Relationships can completely reverse, or go from strong relationships to negligible ones and vice versa. Some correlations can even improve as participants start paying more attention to some relationships that had not been striking enough to catch their attention during normal times.

The “asset diversification” theory suggests that opposite and stable relationships make asset classes ideal candidates to be included in a portfolio aiming for minimal risk at minimal cost. Yet many portfolio managers who followed this doctrine, including some renowned value investors, failed to notice as these correlations broke down during the current crisis and were unable to avert huge losses to the tune of 25-30% (Sargen et al 2009).

For policymakers, it is just as important to figure out the likely implications of policymaking on inter-asset-class linkages during a financial crisis as it is during normal-to-boom periods. Indeed, the effect of fickle inter-asset linkages on all market participants with their capital spread across different asset classes – physical or financial – add up to a major impact on an economy.

It is also critical that nations manage their economic policies and spending in coordination with markets so that market-driven prices do not push them into a sovereign debt crisis.

Understanding inter-asset-class correlations during good times and bad

An analysis of inter-asset-class correlations helps us understand the changing undercurrents between different asset classes (bonds, commodities, forex, and equities being the four most widely participated among them) during the ongoing global financial crisis (see Table 1).

While we chose the 25-year-long stint from 1983 to 2007 as a normal-to-boom period because the economic climate during that period was mostly stable, except for the turbulence during the East Asian currency crisis and the dot.com bust. We have analysed relationships for the entire period to find out the impact of correlation breakdown (during the crisis) on long-term portfolio, which the “value investors” create and maintain.

It is worth noting that during the major part of the pre-crisis period, i.e. 1980s and 1990s, financial innovations had not been as vibrant and frequent as they had been during the first decade of the 21st century in the run-up to the current crisis. Access to markets had been limited, information flow had not been as fast as it is currently, and adoption of technology in trading, clearing, and settlement processes had not been as widespread except in the spot markets for currency.

Results of our correlation analysis (see Table 1) clearly show that currencies and bonds were the most fickle asset classes during crisis. Unlike in the pre-crisis period, their correlations with other asset classes during the crisis period show a weakening, a strengthening, or a complete reversal. Investors’ rationale seems to weaken or simply disappear during a crisis owing to their changing behaviour. Further, the risk-taking ability that largely governs inter-asset relationships during a pre-crisis period seems to wither in the risk-opacity of a crisis.

Table 1. Inter-asset correlations: pre- and post-financial crisis ’08-‘09

1983-2007 – Before the current financial crisis
Correlations
Bond
Commodity
FOREX
Equity
Bond
1.00
0.20
-0.67
0.62
Commodity
0.20
1.00
-0.32
0.57
FOREX
-0.67
-0.32
1.00
-0.31
Equity
0.62
0.57
-0.31
1.00
 
2008-2009 (2 years) – During the financial crisis
Correlations
Bond
Commodity
FOREX
Equity
Bond
1.00
-0.70
0.57
-0.62
Commodity
-0.70
1.00
-0.84
0.91
FOREX
0.57
-0.84
1.00
-0.88
Equity
-0.62
0.91
-0.88
1.00
 
1983-2009 – The overall period
Correlations
Bond
Commodity
FOREX
Equity
Bond
1.00
0.34
-0.70
0.63
Commodity
0.34
1.00
-0.44
0.56
FOREX
-0.70
-0.44
1.00
-0.37
Equity
0.63
0.53
-0.37
1.00

Note: Instruments in different markets used for analysis: Bond market - 10-Year US Treasury Note; Commodity market - SPGSCI commodity index; FOREX market – Dollar Index; Equity market – S&P 500; Data source: Bloomberg

“Unusual” is “usual” during a crisis

Participants’ behaviour (to a large extent), the varying attractiveness of different asset classes, and the risk environment (to a lesser extent), are some of the key reasons behind “unusual” becoming “usual” during a crisis.

Normally, bonds have an inverse relationship with interest rates; higher bond prices equate to lower interest rates and lower bond prices equate to higher interest rates (Norris 2009). And based on this relationship treasuries too have an inverse relationship with the dollar; a rise in US interest rates increases demand for dollars and decreases dollar supplies in the forex markets and vice versa, as figures in the table corroborate, for the normal period of 1983-2007. This is ‘usual’.

What happened during 2008 and 2009 was the opposite. Markets were infected with uncertainty and tight credit while long-term interest rates were expected to rise as investors were less confident of getting their money back. Yet even when uncertainty continued to increase, US Treasury bond prices soared and their yields plummeted (Fennell 2010).

This odd turn of events was largely fuelled by the flooding of markets with cash following the US Federal Reserve’s policy of monetary easing and the US Treasury’s printing of more money. As a result, inflation fears started creeping into the minds of traders and treasuries emerged as the safest investment against inflation. Non-US investors’ demand for dollars in lieu of US treasuries also increased. Thus, during the financial crisis traders and investors transferred their funds from what they perceived to be risky assets, such as stocks, selected currencies, and commodities, to investments they perceived to be safer, namely, US treasuries and the dollar. This led to an “unusual” direct relationship between US treasuries and the dollar index during the crisis period (see Table 1). Moreover, the rising correlation indicates that investors ignored the economic rationale behind the traditional relationship between bonds and currencies and reacted to day-to-day signals on the economy, leading to a fickle relationship between the two asset classes (Kisling 2010).

Other unusual relationships

Considering other relationships, the usual scenario is that, as bond prices fall, so do stock prices – albeit with a lag. The logic behind this relationship is that inflation makes business operations costlier. It is logical that during such times companies and, therefore, their shares will not do so well (Mitchel 2009).

Changes in the bond-stock relation have large implications for investor portfolio allocation between these two asset classes and determination of diversification benefits (MSCI Barra Research Bulletin 2009). In the current crisis, the traditional positive correlation became negative, as the investors kept moving out of the perceived riskier assets including stocks into safer asset classes such as bonds – despite the Fed’s move to slash interest rates aggressively.

It is widely acknowledged that stock markets usually increase or decrease largely in line with corporate performance. On the other hand, the forex market, macroeconomic in nature, moves in line with economic and political trends. Again, as stock market sentiments have a bearing on the overall economic environment they thus influence the forex market to some extent. This explains why, in a normal period, the stock market has a weak correlation with the forex market. During the crisis of 2008-2009, as the corporate sector slid into a hole, the stock markets tanked. But the forex market (dollar) firmed up as it emerged as one of the safest asset classes amid uncertainties compared with other investment avenues. Thus, during the crisis, the usual weak correlation between the stock and forex markets turned into an unusually high correlation.

Commodities also demonstrated a drift from their traditional relationship with other asset classes during the recent crisis (Danske Markets Report 2010). Unlike 1983 to 2007, commodities had a better correlation with the stock market during 2008 to 2009. As stocks fell during the crisis, commodity prices also declined due to a fall in economic performance. The weak positive correlation between commodities and bonds during the normal-to-boom period that is a puzzle to the common economic logic becomes a strongly negative correlation, well-explained by the economic relationship between the cost of capital and commodity prices.

Inter-asset relationships – fickle but dynamic

Markets today represent an interconnected network (Lien 2009) of global participants wherein, given the vastness, the currency market – and, to an extent, the bond market – plays a major role in influencing prices and hence the inter-asset-class relationships – as seen from the above discussion. It is widely known that government policies have a major bearing on the bond and currency markets and hence their impact on inter-asset linkages. Further, though it is seen that bond and forex markets influence commodity and stock markets during normal-to-boom periods, the reverse may also be possible during a crisis. The intensity of changes in such relationships tends to deviate during a crisis not only impacting the savings of individuals and corporations but also the sovereign wealth and debt.

As discussed above, the unusual relationships between asset classes during a crisis are in large part the result of a switch in market participants’ behaviour from applying economic logic, to an altogether different mode of thought in response to the frequently changing government policies unfolding in front them. At such times, it makes sense for market participants chasing returns to keep trackof several often-ignored risks that stem from these fickle-yet-dynamic relationships between assets. At the same time, governments (including central banks) must ensure their policies provide a strong signal to the market and that they pick up the right signals from the market themselves.

This column reflects the personal views of the author.
References

Dankse Markets Report (2010), “Commodities 2010 - Five themes to drive commodity markets this year”.

Fennell, Aaron (2010), “Trading Bonds During Debt Crises”, let’stalkfutures.com.

JP Morgan Asset Management (2009), “Non Normality of Market Returns - A Framework for Asset Allocation Decision Making”.

Kisling, Whitney and Elizabeth Stanton (2010), “Fear Feeding Greed with S&P 500 Correlation to Bonds at Record”, Bloomberg.com.

Lien, Kathy (2010), Bond Spreads: A Leading Indicator for Forex”, investorwords.com.

Mitchel, Cory (2009), Intermarket Relationships: Following The Cycle”, investopedia.com

MSCI Barra Research Bulletin (2009), “The Stock-Bond Relationship and asset allocation”.

Norris, Jay (2009), “Bonds Point to Higher Rates in States”, trading-u.com.

Sargen, Nicholas P, Thomas L Finn, and Charles A Ulbricht (2009), Reflections on the Financial Crisis: When Diversification Failed”,

Wilson, Richard C (2010), “Funds of Hedge Funds Manage to Survive Financial Crisis”, hedgefundblogger.com. 

420 Reads