Are ETFs boosting correlations?

I came across a recent study that points out to a sharp rise in correlations between stocks right about the time when the ETF market has experienced exponential growth. Is it possible that the popularity of ETFs has led to a jump in equity correlations? It does seem plausible if we consider the way ETFs trade.

When you want to purchase an ETF, you usually go to the secondary market where buyers and sellers congregate. When the purchase amount is significant, however, it needs to be created through the primary market and the way this is done is through one of the designated brokers that deal directly with the sponsor in the primary market. The broker uses the investment cash to purchase a basket of stocks that reflect all the constituents and their respective weights of the underlying index. This is then exchanged against an equivalent number of units of the ETF. In the event of a redemption, the exact opposite happens, whereby the broker exchanges units of the ETF for an equivalent basket of stocks representative of the index. Now imagine a bear market scenario where a whole lot of redemption activity is going on. In the case of a broad market index ETF, the broker will have to liquidate a large number of stocks right about the same time. In other words, all the constituents of the index will be experiencing selling pressure at around the same time which means an increase in the correlation between them.

This estimated general increase in correlations has led to a situation where you need to hold a larger number of stocks to achieve an optimal level of diversification. It also further emphasizes the importance of ensuring a broad level of diversification between different asset classes.

Just about right...

The Fed has kept monetary policy aggressively loose and intends to do at least until sometime next year. Of course, this will depend on how the economy unfolds. Loose monetary policy has been the necessary medicine to keep the economy afloat, ever since the housing market bubble popped in 2007 but it has largely been ineffective in part because the Fed found itself in a "liquidity trap". The transmission system between Fed policy and bank lending is broken but there are signs that it is healing.

Monetary policy, in "normal" times can be a potent tool, which is why it needs to be wielded carefully if price stability is to be maintained (one of the two mandates of the Federal Reserve). One tool that can help determine whether the target rate is set "correctly" is what is known as the "Taylor Rule" named after the U.S. economist John Taylor. In its basic form, the Taylor Rule is a linear equation that determines what the Fed target rate should be. To do this, it takes the current rate of inflation and compares it to the "optimal" rate of inflation. It does the same with growth by comparing current output or GDP to "potential" GDP. Basically, the required target rate changes if either of these two factors are currently above or below their optimal or potential levels.

The graph above compares the actual Fed Funds Rate (white line) to that of the Taylor Rule Estimate (blue line). It is interesting to note that overall, the actual target rate and its estimate seem to be relatively well correlated and that currently the Taylor rule estimate seems to agree almost fully with the target rate. In other words, the estimate seems to fully support Fed policy.