"Bubble" is an emotive term used to grab the attention of the nervous investor. It conjures up memories of the 2008, 2001 or 1987 stockmarket crashes. But the term is too often used without explaining to the reader what a bubble in other markets such as bonds actually means
We look at the last 100 years to show the worst corrections in equities, bonds and property markets, and show a bubble correcting in equities is likely to cause 3 to 4 times more pain than one in bonds, with housing somewhere in between
When you hear that a “bubble” is about to burst in financial markets, how much of a loss does that make you think of? 5%? 10%? 30%? Problem is there is no definition, but it’s associated with painful experiences such as the technology stocks bubble in 2001 or the US housing market bubble that brought down the world in 2008.
And that’s the problem. “Bubble” has become a term used in financial media more and more to invoke thoughts of losses of the magnitude of those market collapses, for example 40-50%, but used in association with markets without the volatility to ever fall anywhere near that far such as bonds. Headlines sell, but there is no data in the articles to give the reader a sense of what a bubble actually means and what the losses from the bubble bursting might be.
If commentators imply that something terrible is about to happen, they should let people know the implications.
So here’s a translation guide to help next time you hear that term. Then you can have the more subjective debate about whether there is a bubble or not and understand why it matters.
What is a bubble?
Firstly, the term itself comes from the fact that when assets go up in value we say that they are “inflating”. Inflating asset prices are a good thing – that and income are the reasons we invest. If assets inflate beyond their fair value, that is they are over-inflated, and that continues for some time, we call that a bubble. What typically happens when a bubble bursts is that prices fall back to normal levels or even beyond. So any asset market can be in a bubble if it is over-inflated for an extended period of time, but the bursting of the bubble and fall back to normal levels results in very different loss scenarios for investors. Riskier markets tend to over-inflate much more than lower risk markets.
The problem with a bubble is that it’s hard to know you are in one until the crash comes. So the key is to know what the implications of the crash could look like and how you would cope if this occurred. The below table shows some historic data to provide relativities between the way different markets have fallen in the past. The obvious, yet seldom reported conclusion is that bonds don’t fall as hard as riskier asset classes like equities or housing. It’s obvious because they also don’t rise as much, they are simply less volatile.
Worst crashes in the last 100yrs
The worst correction in Bonds was one sixth of the size of the worst equities correction.
Source: Federal Reserve Economic Database; Dr Robert Shiller
Note: The last column details the expected return assuming the portfolio is equally split between all three asset classes and held for the years listed in the first column. We assume interest is compounded annually. In 2009 the portfolio is just held for the single year.
Worst years over the last 100 years
The worst year in bonds would not make it on to the worst 10 equities years, and the average of the worst ten years for equities is more than a massive 4 times worse than that for bonds. Bonds are lower risk and show lower volatility.
Source: Federal Reserve Economic Database; Dr Robert Shiller
Worst months over the last 35 years (using Corporate Bonds and Equities* this time)
Even the “infamous” bond crash of 1994 resulted in losses that would have ranked outside the worst 50 months if compared to equities. Corporate bonds are higher than risk than government bonds, but the worst equities months are still 3 times worse than the worst corporate bond months.
Source: Federal Reserve Economic Database
*Using Corporate Bonds is more relevant than Government Bonds used in the first two tables, but there is not reliable data going back further than 1979.