This week’s chart shows the dynamic nature of correlations between asset classes by comparing correlations amongst traditional asset classes over 20-year¹ and 5-year² historical periods.³ The chart above shows how much these correlations have all increased when comparing the 5-year figures to the longer dated 20-year period. What does this mean for investors? We see two main takeaways:

1. For those that rely on mean-variance optimization programs for determining their asset allocation, it is imperative to understand the exact timeframe reflected in the correlations used as inputs, as different time periods will yield not only different correlations but critically, different portfolio structures.
2. The correlation amongst traditional asset classes has increased in the last five years, thus it is more difficult to truly create a “diversified” portfolio that offers protection from large draw downs in the equity markets. This was never more apparent than during the financial crisis of 2008-2009.

As now outlined in both this chart and “Correlation Doesn’t Tell the Whole Story”, correlations can be helpful in conducting asset allocation studies, but they also feature some notable shortcomings that should be well understood by those who rely on them for portfolio decisions.

¹ March 1992 – February 2012
² March 2007 – February 2012
³ Indices used for analysis were Russell 1000, Russell 2000, MSCI EAFE, MSCI Emerging Markets, and BarCap Aggregate

This week’s Chart of the Week covers increased correlations among asset classes. Correlation is a statistical measure showing how two variables move in relation to each other. It can range anywhere from -1 to1, and refers only to the direction of changes. Perfect negative correlation (-1), implies that two items move in completely opposite directions and perfect positive correlation (+1), implies that two items move in lock-step. The chart above shows the five-year correlations between the S&P 500 and a broad set of sample asset classes. As evident, all risk-assets have had a very high positive correlation to the S&P 500 over the past five years. Referring to the chart, aside from the three Treasury indices, no other asset class has a correlation less than .90, which is extremely high.

However, correlation only tells part of the story: just because two asset classes have high correlation does not mean that their returns will end up being the same. In fact, this will most likely never be the case. For example, over the time-frame captured above, despite very high correlations, emerging markets, the S&P 500, and the REIT index have annualized returns of 2.8%, -.56% and -3.48% respectively. Over the past five years, as globalization has become more pronounced and economies more intertwined, correlations have certainly increased to all time high levels. But since correlation does not capture magnitude of returns, investors should continue to utilize an asset allocation model that takes potential risk and return into account.

On Tuesday, the ISM factory index for December was released, with last month’s level reaching 53.9, the highest since April. Perhaps more importantly, this was above expectations of 53.5, thus providing an unexpected surprise to the upside to kick off 2012. Stocks and commodities rose in reaction to the better than expected data, and while the economy is far from back to full health, this was welcome news to kick off the 2012 year in the financial markets.

Given the fourth quarter U.S. stock market performance to date, we have been asked if certain quarters have historically offered more positive performance. Based on S&P 500 data from 1926 through 3Q2011, the answer seems to be yes, as there does appear to be some persistency across the four quarters. As the graph demonstrates, the fourth quarter has offered the highest frequency of positive returns and highest average return (3.55%). For the sake of comparison, the S&P 500 is up about 8.6% through October 25th, so while we would expect some reversion to the mean as the year winds down, there is reason for optimism about this quarter’s stock market returns.

This week’s COW takes a look at the Volatility Index (“VIX”), defined by the CBOE as the measure of short-term stock market volatility conveyed by S&P 500 option prices. It is also known as the “markets fear index”, as VIX tends to rise when markets are falling. Although the VIX has been extremely volatile since the Financial Crisis of 2008, we chronicle the events of the last two months in an effort to further illustrate the dramatic equity market movements of summer 2011.

Looking at the chart, we first notice the overall inverse relationship between the S&P 500 index (red line) and VIX (gray line); when one index is falling, the other is rising – not surprising, since we would expect market fear (as measured by the VIX) to increase when the equity market (S&P 500) is falling. Second, the month of July was relatively quiet, as neither index showed much movement over the course of the month. However, as August arrived, several events triggered substantial movements in the two indices. We focus on three of the most notable:

• On Monday, August 8^th, S&P downgraded the United States’ credit rating from AAA to AA+; VIX saw a 50% intraday gain from 32 to 48.

• On Thursday, August 18^th, VIX closed 35% higher than the previous day in the wake of more rumored problems for European banks, settling at 42.67 by end of day.

• Finally, August 24^th featured another large movement in the VIX index when it was announced that the CEO of Apple, Steve Jobs, was resigning due to health concerns. VIX quickly subsided though as markets expressed confidence in his successor to maintain Apple’s impressive run.

For the sake of comparison, the five-year average of the VIX index is 24.32; thus these elevated figures in August certainly reflect a higher than normal volatility, which has indeed played out in the equity markets. Although the figures do not approach the all-time high of 96.4 when markets were collapsing in October of 2008, the elevated levels have made investors stand up and take notice. Unfortunately, the VIX will likely continue to be volatile, which is a direct reflection of expected choppiness in the equity markets.

This week’s chart looks at the amount of excess reserves banks are holding at the Federal Reserve (orange line) along with the corresponding changes to the Federal Reserve’s balance sheet (black line). As of the end of July 2011, banks are holding over $1.6 trillion in excess reserves, which is notably higher than what historical averages would suggest. This has led some market commentators to worry about inflation escalating as banks begin to lend out those assets (note that overall loans and leases issued by commercial banks, as represented by the red line, have fallen since the Financial Crisis of 2008 – 2009).

In the United States, all depository institutions (commercial banks, savings and loans, credit unions, etc.) are required by the Fed to satisfy a reserve requirement (or liquidity ratio): each bank must retain a certain amount of cash that cannot be used for loans to business or consumers. Any reserves held above the required amount are considered “excess reserves”.

Historically, institutions have minimized their excess reserves because it is generally more profitable to issue loans than to hold the cash (the banks can earn higher amounts of interest via loans). When money is lent by banks, it can have what is called a multiplier effect on money supply: because banks only have to hold a small percentage of assets to meet the reserve requirement, they are able to increase the money supply by lending. For instance, with a 10% reserve requirement, $100 could theoretically be turned into $1,000 through the multiplier effect.

As a result of this theory, many market commentators have expressed concern over the current large level of excess reserves held by banks. Based on this macroeconomic theory, even a relatively small decline in excess reserves of $100 billion could theoretically increase the money supply by $1 trillion (assuming a reserve requirement of 10%).

The Fed is confident that inflation is not a near term threat to the economy, and when (if) inflation does become a concern, it has the tools to control the growth of price levels. A large output gap in the U.S. economy, high levels of unemployment, and slow wage growth are the most commonly cited reasons for the expected low levels of future inflation. The Fed also has a few tools to help combat inflation from excess reserves, including selling large amounts of securities from its balance sheet (either outright sales or using reverse repurchase agreements) or increasing the interest rate it pays on excess reserves to incent banks to continue to hold excess reserves. The ability to pay interest on excess reserves is a new tool for the Fed and in its view alters the theory behind the multiplier effect.

The fact remains that the U.S. is in uncharted monetary policy territory, with little historical precedent to rely on. As Fed Vice Chairman Donald Kohn said, “the calibration of our exit from these policies is complicated by a paucity of evidence on how unconventional policies work. We will need to be flexible and adjust as we gain experience”.

This week’s chart examines the frequency and magnitude of market corrections in the U.S. equity market, as measured by the S&P 500 Index. A market correction is defined as a decrease of 10% or more within one calendar year. Using data back to 1950, we found that every year featured at least one market drawdown, and over half of those years (35 of the 62 years, approximately 56%) were true market corrections. What is even more interesting is how large some of these corrections were, with 11 of those years seeing intra-year declines of over 20%. So while the steep drop over the last week has contributed to an 18% decline in the S&P 500 Index (through Monday’s close), perhaps investors can find some reassurance knowing that more severe market corrections have occurred in the past.

Last week’s chart addressed the increase in IPOs during 2011. In addition to the number of companies coming to market, the returns of these companies post-offering can also serve as an important metric. So far in 2011, stocks that have been public for less than one year, as measured by the Bloomberg IPO index, have returned a positive 4.1%, yet are lagging behind the broad Russell 3000 index which is up 6.7% for the year. While the returns of newly listed companies are often linked to the general direction of the market in which they trade, severe dislocations like the occurrence in 1999 are cause for concern. Although there is some concern about the rapid price increase witnessed for some of the IPOs in 2011, we are still far from the misplaced exuberance of 1999.

This week’s charts show the cumulative outperformance of the two Fama-French Factors, SMB and HML. SMB stands for small minus big, and is the excess performance of small market cap stocks minus large market stocks controlling for value. HML stands for high minus low, and is the excess performance of high book to market stocks minus low book to market stocks controlling for size. Graphs are shown on a log scale to highlight relative change. A change in relative size on a log scale chart is the same visual distance, which enables the viewer to more easily compare different fluctuations over time. For example, on an absolute scale graph, a price change from 50 to 100 appears much smaller than a price change from 100 to 200, even though both represent a doubling in price. On a log scale graph, these two changes appear the same.

As shown on the graphs there have historically been excess returns to small stocks and value stocks over the long term. While the volatility of both of these factors has been high (11.53 for SMB and 12.41 for HML), the value factor has earned a much higher premium over time than the size factor. This has led to long periods during which small cap stocks underperform large cap stocks, as highlighted on the graph.

While it is impossible to say given the data here, it is certainly plausible that there are business cycles that are more or less favorable to small companies compared to large companies. In fact, some academic research suggests small companies are more susceptible to shocks in profitability, which explains some of their underperformance starting in the mid 80’s. See Hou and van Dijk (2010) for more details.

The takeaway for investors is that while small caps may outperform over the long term (30 years or more), there can be extended periods during which they underperform large cap stocks. Small caps have certainly been a good bet over the past decade, but their strong performance versus large caps may or may not continue for the next.