Friday, January 13, 2017

Bond Market Volatility - Lessons from the Past

A rather fascinating article posted on the Bank of England's Bank Underground website by Paul Schmelzing at Harvard University takes a very, very long look at the global bond market.  In fact, given that his data set goes back to 1285, this is one of the longest long-term analyses of the blond market that I've seen anywhere!  Additionally, the author provides us with a glimpse of what may lie ahead for the bond market by looking at past bond market "readjustments".

Let's open with this graphic from his posting which shows the global risk free interest rate since 1285 along with the bond bull markets that are shaded in blue:


As you can see, at 36 years, the current bond bull market was one of the longest in history with only two previous bull markets being longer; the bond market rally at the peak of Venice's commercial domination in the mid- to late-1400s and the century following the Peace of Cateau-Cambresis between 1558 and 1664.

Here is a graphic showing the length and size (as measured using the fall in interest rates) of bull markets since 1285:


There is only one bond bull market that saw interest rates fall by more than the current bond bull market (between 1981 and the present) and that was back in 1441 to 1482.

Let's take a closer look at bond bear markets.  According to the author, there are three types of bond bear markets as follows:

1.) Inflation Reversal:  This type of bond bear market was experienced in the years between 1967 and 1971 and is associated with a sharp change in consumer price inflation (CPI).  Between the years of 1965 and 1970, annual increases in CPI tripled from 1.6 percent to 5.9 percent; this resulted in U.S. bonds losing 36 percent of their value in real terms.  Here is a graphic showing what happened to bond price returns (dark blue bars) and the yield on ten year Treasuries (light blue line):


2.) Sharp Reversal:  This type of bond bear market was experienced in 1994 and is commonly known as the "The Great Bond Market Massacre".   This is what happened to ten year Treasury yields:


Ten year yields jumped from 5.7 percent in January 1994 to 7.49 percent in May 1994.  This is conventionally blamed on the Federal Reserve which tightened short-term rates by 2.25 percentage points over the entire year in its concern about inflationary pressures.  The increase in interest rates resulted in bond market losses of more than $1 trillion.  As Fortune magazine noted back in 1994, the reason for "...such staggering losses is the sheer size of today's bond market...".  Given that the total public debt (including both debt held by the public and intragovernmental debt) was only $4.536 trillion in January 1994 which has mushroomed to $19.977 trillion with $14.435 trillion of that being held by the public, we can see that a sharp reversal scenario today could result in catastrophic losses.

3.) The VaR (Value at Risk) Shock:  This type of bond bear market was experienced in 2003 in Japan when actions by the Bank of Japan resulted in a yield curve steepening as shown on this graphic:


Prior to March 2003, there was a significant flattening of the yields of Japanese Government Bonds (JGB).  In mid-2003, there was a rapid steeping of the JGB curve, thanks in part to actions by the Federal Reserve and changes in the expected duration of the Bank of Japan's quantitative easing program.  This meant that, in order to reduce their risk profile on their holdings of domestic government bonds (i.e. the Value at Risk), Japan's banking sector had to sell their JGB assets, which resulted in an interesting conundrum; the more bonds that the banking sector sold, the more the price dropped and the more the yields rose.  This forced the banks to use interest-rate swaps as a hedging tool to reduce further losses on their bond portfolios.   It is interesting to note that this occurred after the Bank of Japan announced its experimental quantitative easing policy in March 2001 in a desperate attempt to prod some life into the Japanese economy.         

The author of the article suggests that a repeat of the 1994 bond market massacre is rather unlikely, however, factors including a move by central bankers to allow inflation to overshoot their current targets could create a situation where an inflation reversal-type bond bear market could occur.  As well, the VaR risks in the banking sector have increased, particularly in Japan and Europe which could lead to a value at risk bond market shock, particularly since banks in both regions have a "home bias", preferring to hold bonds from their home nation.  In addition, recent studies have suggested that bond market liquidity problems could occur, resulting in bond sellers finding far fewer buyers for their wares than they expect particularly during a crisis, an issue that could put further downward pressure on bond prices.

Let's close this posting by quoting from the author's final paragraph:


"On balance, then, more than to a 1994-style meltdown, fixed income assets seem about to be confronted with dynamics similar to the second half of the 1960s, coupled with complications of a 2003-style curve steepening. By historical standards, this implies sustained double-digit losses on bond holdings, subpar growth in developed markets, and balance sheet risks for banking systems with a large home bias."

3 comments:

  1. History confirms your reason for concern. A read of the pdf file of the 2009 bestselling book titled, "This Time Is Different" did little to convince me that this time is different.

    It chronicles eight centuries of financial follies in which financial meltdowns have typically followed real-estate bubbles, rising indebtedness, and gaping deficits. More on why many of us see a strong similarity between what is happening today and prior financial meltdowns that have resulted in crisis can be found below.

    http://brucewilds.blogspot.com/2016/12/thoughts-on-this-time-is-different.html

    ReplyDelete
  2. Great post. I would like to link to this and possibly use one of your charts. Do you have an attribution policy and, if so, what might it be?

    Thanks very much

    ReplyDelete
    Replies
    1. I have no problem if you want to link to this article. The graphics used are from the original article on the Bank of England's website.

      Thanks for reading.

      Delete