Bond Yield Curves and Media

As I woke up today, I started by reading this article with the BBC which is arguing that because we are seeing negative interest rates that we are headed for a financial crisis. The issue with this is that the real signal for economic downturns is a negative yield curve, such as when your 30 year bond gives you a lower return than your 15 year bond (for example). It is not when your bond yield is below 0. All that I see the market return is that demand for bonds is too high, and the demand is so large right now the cost of holding bonds rises to the point past our arbitrary 0 lower bound which turns out to have been a fundamentally wrong assumption. The amazing thing is that even though bonds literally cost lenders money (those who buy bonds from governments, companies, etc.) people are still going to bonds.

This is uncharted territory. It is impossible for economists to know what this means because it has never happened before. It is possible that the lack of faith in markets is a sign of a weak economy in the near future, or it is just due to a shift in the supply and demand for bonds. If governments issued more bonds we would see interest rates rise, and given how all of Europe is gripped in an anti-debt paranoia currently with their right wing governments this doesn't seem unreasonable to me. Greek debt has stabilized over the last few years, Japan's debt/GDP ratio grew by only 3% last year, the US is in the grips of people in Congress who are on an anti-debt crusade, Germany's deficit is almost nothing, so it looks a lot to me like we have a reduced supply of bonds which will make them more expensive. Also, with inflation as low as today it means a negative yield curve is not as big of a deal as it would have been in 1982.

I doubt that it is due to an increase in demand for bonds fleeing to safety from a rough economy about to go to Armageddon because we have never seen people flee to bonds for safety to the point where they reach such low rates of inflation before.

Another thing is looking at the long term, it seems like we have been in a period of decreasing bond yield for 30 years. Whether one looks at Germany, The United Kingdom, Japan, France, Canada, Australia, or the United States, one find the same pattern of unusually high interest rates in the early 1980s and a continual decline since then back to what are actually more normal nominal bond yield prices. We have also seen the rate of inflation go down significantly from the early 1980s, meaning that it seems like bonds change their prices slower than the economy as a whole. We might have been watching 30 years of recovery from what a very volatile period with record unemployment and record inflation of the 1981-1985 period. Getting back to no significant return on bonds might be simply a return to the old normal. The change in the bond yield price also doesn't appear to me to be in line with recessions, and mostly random in the short term rises and declines. This current drop in bond prices could be part of the long term trend of volatility. This time it just hit a negative rate which makes people freak out because we haven't seen US bond prices below 4% in over 50 years.

Also, there is a lot of misinformation rolling out, first from the BBC:

There is also a potential problem about large losses being incurred by the funds that own bonds with negative yields. The problem is that the price of a bond and the yield move in opposite directions. So if those negative yields turned sharply upwards, above zero, their prices would fall and the bondholders would lose.

The author is correct that the price of a bond inverse of the yield curve. So a higher yield curve is a lower price to hold a bond (price in this case is best seen as opportunity costs) and he is right that the price of bonds will fall if yields go up. However, he puts a very peculiar and completely incorrect spin at the end of this where he says bondholders will lose. He is arguing that if bondholders get a better return for their money they are losing. This is backwards and incorrect. Bondholders would like to see higher interest rates for their money, which seems obvious to me, but apparently the economic editors of the BBC want to get a lower return for their money. Given their coverage of the austerity crisis, their claimed imminent secession of Scotland that dominated their website last year, and the little BREXIT that couldn't (given betting odds and the polls) I'm not surprised they will let something as completely wrong as this statement get out. I'm going to delete the app from my phone for such reliable ludicrousy, though given the fun of debunking such claims I might keep them. TBD

On the same lines in economic lala land, The Wall Street Journal by our favorite hate mogul Rupert Murdoch has an equally crazy statement.

The narrowing gap between short-term and long-term rates may not be the signal it once was, but the Federal Reserve should still pay heed to it

The first issue with this mess is that we have no proof or evidence that negative yield curves are no longer a signal, and looking in history all negative yield curves have been followed by recession or are currently negative. They are ignoring hundreds of years of research in this statement and a magazine which claims to be about finance should no better than this. The editors of the Wall Street Journal obviously don't know much about economics or are lying to their readers.

Stick to the fundamentals, stick to what we know. Do not make big assumptions that "this is the big one" or that all historical knowledge of what works is now invalid because aliens or whatever. People who go assuming this time is different every six months tend to be wrong and people who stick with the basic economic theory have a very annoying habit of being right in our predictions. I do not see evidence that we are looking at an economic recession, however I do see evidence in a decline of available bonds which raises the price of government debt and a continuing 35 year long trajectory of declining bond prices which we are still observing.