The chart above depicts the amount of Italian government bonds held by Italian banks. As seen in the chart, this amount hovered between $200 and $250 billion until December of 2011, which is when the first round of the Long-Term Refinancing Operation (LTRO) began. A notable aspect of the LTRO was that it allowed Italian banks to borrow funds from the ECB for 3 years at a 1% interest rate with appropriate collateral. Given these favorable terms, Italian banks used portions of the funds obtained from the ECB to invest in higher yielding bonds issued by their government. As a result, the number of Italian bonds owned by Italian banks has since swelled to nearly $350 billion. Ultimately, this trend has been helpful to lower the interest rate on Italian bonds, but is troublesome since a large portion of the debt that Italy issued has been purchased by Italian banks because foreigners have been skittish to invest. Foreign investors have cut their holdings of Italian governments bonds to their lowest level since 2005. As Italian banks purchase more Italian government bonds, the public and private sectors become more intertwined.

Italy has a substantial portion of debt maturing in the next few months (table below). It remains to be seen if there will be sufficient demand (foreign and domestic) to refinance these obligations at a reasonable interest rate.

Italy – Principal Maturity

ECB president Mario Draghi has indicated a willingness on behalf of the ECB to step in and purchase Italian bonds that fall on the short end of the yield curve to ensure liquidity. This implied guarantee means that Italy will most likely issue short term debt to attract investors. By issuing short-term debt, Italy essentially buys itself a little more time to demonstrate economic improvement to attract foreign investment. However, a lack of progress will likely translate into larger near-term debt and rising interest rates.

Most investors understand that over the long-term, the investment grade credit sector tends to outperform the U.S. government sector due to its risk. Many may be surprised however, at the size of the outperformance. Since its 1973 inception to May 2012, the BarCap U.S. Credit Index has outperformed the BarCap Government Index by just 38bps. Today’s Chart of the Week shows the rolling three year outperformance of the BarCap Credit Index versus the BarCap Government Index. Up until 2007, credit enjoyed a long period of relatively steady outperformance, interrupted by the recession after the tech bubble burst.

To achieve this outperformance, investors in credit would have experienced a standard deviation of 7.79%, compared to 5.61% for governments. The maximum decline for credit versus governments was -19.26%, and -7.92% respectively. This decline in credit was actually experienced from September 1979 to March of 1980, not in 2008.

Thus, while the credit crunch of 2008 was an extreme event, the actual margin of underperformance in credit had been seen before in the 1980’s. It is also important to note that the raw outperformance numbers are likely overstated in favor of the credit index, which has a longer duration. As rates have been in a secular downward trend over the past 30-years, this has boosted longer duration returns. The duration of the BarCap Credit Index as of May 31 was 6.73, compared to 5.43 for the BarCap Government Index. From 1997 to 2012, the period for which duration numbers are available, the credit index has had an average duration 0.91 higher than the government index.

Of course, past returns do not necessarily predict future returns. The forward outlook for credit has tended to be more positive during periods of elevated spreads, as investors are paid a higher premium to take on credit risk. As of the end of May, the option adjusted spread of the BarCap Credit Index was 193bps, compared to its long-term average of 121bps.

Much has been made of the low bond yields on risk-free U.S. government debt. The yield on 10-year government debt dropped to just 1.7% as of May 18th. Many investors assume that if rates stay where they are today bond returns will be less than 2% over the coming years. However, this ignores both the steepness of the yield curve and the fact that most institutional investors maintain a fairly constant duration in their bond portfolios. This chart shows the current yield curve (red line) and the expected return an investor can expect to achieve owning government bonds at each maturity along the curve, assuming they maintain a constant duration. As the chart shows, investors in 10-year bonds will earn almost double the 1.7% return indicated by the yield curve if rates remain unchanged.

While this may come as a surprise, the math is fairly straightforward. An investor that buys a 10-year U.S. government bond today will pay $100.47 and receive a $1.75 coupon. The 10-year risk free interest rate is 1.70% and the 9-year risk free rate is 1.51%. This means that one year from now our investor owns a 9-year bond that pays a $1.75 coupon. However, because the current market interest rate for a 9-year risk free bond is 1.51%, the bond has appreciated to $101.99. To maintain the duration of the investment, our investor sells the 9-year bond at $101.99 and buys a new 10-year bond. Since rates have not changed a 10-year bond still sells for $100.47. Our investor’s total return is thus:

($101.99+$1.75)/$100.47 = 3.25%

This simple illustration and the implications for an institutional portfolio are discussed in greater detail in Marquette’s April 2011 Investment Perspectives “Short Duration vs. Core Bonds in a Rising Rate Environment”. Currently, the yield curve is predicting a fairly substantial rise in interest rates a few years in the future. However, if such a rise does not materialize and the current low rate environment persists, bond returns may once again exceed expectations.

Wild uncertainty in the equity markets coupled with European debt concerns have driven U.S. Treasury yields to all-time lows, and left income-driven investors searching for alternative sources of yield. While bonds will always serve as a major component of an income-driven portfolio, the overarching low yield environment has led investors to look beyond the traditional sources of return. This week’s Chart of the Week examines the potential of other asset classes to provide the income streams that have historically been provided by bonds.

As shown above, dividend yields for U.S. large cap equities now exceed that of the 10-year Treasury, and the MSCI EAFE and EAFE Value indices have attractive yields as well, exceeding those of the BarCap Global Bond Index and the BarCap U.S. Corporate Bond Index. The NCREIF Property Index has generated an annual income return of 6.44% as of June, 30, 2011, which provides an opportunity for greater portfolio diversification with minimal correlation. These asset classes are more volatile than traditional bond portfolios in regards to capital appreciation, but the income generation has become favorable to many options in fixed income. While it would be premature to label these trends as a new regime, it will not be surprising to see income-driven investors tilt their portfolios to include greater allocations of higher yielding (and traditionally more volatile) asset classes.

With the August 2nd deadline fast approaching investors are increasingly wondering whether the U.S. might actually default on its debt obligations. In an effort to gain some insight into what the market is expecting, this chart looks at the pricing of Credit Default Swaps (CDS) on U.S. Government debt over the last year. CDS only pay off in the event of a default and are often used by bondholders to hedge against the risk that a bond in their portfolios defaults. As a result rising CDS prices offer a good indication that the market expects a higher probability of default. What is most notable is the jump in CDS prices on 1 Year Treasuries, while at the same time, CDS on 5 and 10 Year Treasuries remain largely unchanged. Given the jump in CDS on shorter duration Treasuries the market seems to be indicating that, if there is a default, it is likely to occur in the next year. However, the lack of movement in CDS prices for 5 and 10 Year Treasuries indicates that the markets are assigning little overall risk of a default (because if a default occurred all Treasury bonds would suffer losses so we would expect to see a jump in spreads for CDS of all maturities). Lastly, to put these numbers in perspective, CDS on 5 year Greek debt trades at 2,320, CDS on 5 year French debt trades at 107, and United Kingdom debt trades at 75. So while there is considerable amount of concern over the debt ceiling deadline in Washington, the markets appear to be pricing in little risk of default.

As many commentators have pointed out, over the past two years the BarCap Aggregate has seen a large increase in its benchmark allocation to treasuries. Since December 2008, treasuries as a percentage of the Agg have grown from 21% to 33% (see first image above – click on thumbnail for larger version). This highlights a drawback of any bond benchmark based on issuance. Issuance based benchmarks by their nature drive allocations not based on expected risk and return, but based on the funding needs of underlying issuers. One of the primary drivers of the increase in treasuries as a percentage of the Agg has been the large federal deficits caused by the 2008 recession necessitating higher treasury issuance. Historical and projected deficits are shown in the second image above (click thumbnail for larger version).

The federal deficit picture shows why the Treasury component of the Agg may stabilize, albeit at a higher level. While the Federal budget remains challenged and the political environment uncertain, through GDP growth alone the Federal deficit will decrease from its peak.

A more important concern for investors is how the change in the composition of the Agg affects the risk and return profile of their fixed income portfolios. If treasuries increased in market weighting in the Agg, clearly other sectors decreased in weighting. The change in sector weighting in the Agg for major sectors is shown below:

Source: Barcap; changes since December 2008

As treasuries have increased as a percentage of the Agg, MBS, agency bonds, and securitized products (ABS and CMBS) have decreased. This is due to a combination of decreases in issuance, especially for securitized products, and Fed purchases (MBS). Notably, the percentage weight to corporate bonds has remained constant on relatively strong issuance, and the weight to credit as a whole has actually increased.

Over the past 10 years, treasuries have been highly correlated with agencies (0.96) and MBS (0.85), and had a lower correlation with credit (0.62). Returns and standard deviations for the past ten years are shown below:

The Taylor rule, proposed by John Taylor, is a formula for determining the target Fed Funds rate. In the Taylor Rule, the Fed Funds rate baseline is set to the target nominal rate (target real rate plus target inflation), and then adjusted based on economic conditions. The rule states that the Fed Funds rate should be raised when inflation is higher than target inflation (“Inflation Gap”), and lowered when economic output is lower than potential output (“Output Gap”). The equation for the Taylor rule is shown below:

Target Fed Funds = Inflation + Target Real Rate + a1(Inflation Gap) + a2(Output Gap)

Though the rule itself is relatively simple, there are many different interpretations of how to implement it. There are of course multiple measures of inflation, and multiple measures of the output gap. In the chart, we use the Core Personal Consumption Expenditures (PCE) deflator, one of the measures favored by the Fed, as our measure of inflation. To measure the output gap, we use the CBO real potential GDP series, a trend line estimate, less actual real GDP.

The other two important components of the Taylor rule are the “weights” placed on the inflation gap and the output gap. These are the terms a1 and a2 in the formula. The larger the weighting, the more the prescribed Fed Funds rate moves in response to changes in inflation and output. In his original formulation, Taylor proposed weights of 0.5 for both inflation and output.

While the Fed does not explicitly follow the Taylor rule, it has proved to be a reasonable approximation of Fed policy. However, the Fed has indicated that it places more “weight” on the output gap than Taylor originally suggested. The graph shows the “original” Taylor rule, as well as an “alternate” Taylor rule with more weight placed on the output gap. In normal times these rules track fairly closely, however, when the output gap is large the alternate rule prescribes a much lower Fed Funds rate than the original rule. This formulation currently suggests a negative Fed Funds rate. Because the Fed Funds rate is up against a zero lower bound, this explains why the Fed engaged in unconventional monetary policy actions such as quantitative easing.

The graph projects out the target Fed Funds rate based on both formulations of the Taylor rule. Real GDP is projected using Bloomberg consensus estimates, and a constant 1.5% inflation rate based on the PCE deflator is assumed. This inflation rate is lower than the 2% target, but higher than the recent reading of 0.9%. Given these assumptions, the alternate rule does not imply an increased Fed Funds rate until 3Q 2012. This is roughly in line with the futures market, which suggests a greater than 50% chance of an increased Fed Funds rate in 2Q 2012.

Finally, this chart is not intended as a forecast, but merely as a template for understanding Fed policy. Any large surprise either to the upside or downside for GDP could impact Fed policy. The more important indicator to watch may be measures of core inflation. The Fed has stated that it favors measures of core inflation (inflation less food and energy), and has described the current commodities led uptick in CPI as “transient.” As long as measures of core inflation remain subdued there is little pressure on the Fed to raise rates until the output gap narrows. Economists should debate why the Fed makes its decisions; investors should only be concerned with how the Fed makes its decisions to determine likely outcomes.

The Fed recently completed its latest stress tests on banks. Based on the results, many banks were given the green light to increase dividend payouts as well as announce share buybacks. With this in mind, our chart of the week looks at the charge off rates and delinquency rates of loans at all commercial banks.

The Federal Reserve calculates these rates based on quarterly reports by all banks. The charge-off rate is defined as the flow of a bank’s net charge-offs during the quarter divided by the average level of loans outstanding. The delinquency rate is the ratio of the dollar amount of a bank’s delinquent loans to the dollar amount of total loans. Loans include real estate, agricultural, commercial & industrial, and consumer.

Prior to the official start of the recession, in the 2nd of quarter 2006, both rates began to increase, serving as a sign of things to come. In 2008, with the financial industry in danger of collapsing, the Fed stepped in and imposed tight restrictions on banks which led to dividends being slashed or all together eliminated. As banks struggled through the crisis, charge-off and delinquency rates climbed through 2009. During the official recession period between the 4th quarter of 2007 and 2nd quarter of 2009, the charge off rate increased by 255% and the delinquency rate increased by 160%. Perhaps as a sign that lending standards have improved and the economy has strengthened, both rates began to fall in 2010 and have been on a consistent decline the last four quarters. However, both rates are still well above pre-recession levels and undoubtedly haunted by continued high unemployment and a struggling housing market.