There is a minimum bet size (you cant buy bonds with cents or low amounts), once you lose a certain amount, you are out of the game and the backtesting ends. Effectively, that's an 100% loss given that whatever is left over will be eroded by inflation With bonds if everything goes well, you make a small return. If things go bad, you lose it all or close to it. With stocks, if everything goes well, you make a bigger return. Compounded over time, that adds up to a lot more than bonds. If things go bad, in stocks you get hit with a huge drawdown that the vast majority of the time, reverses within 10-20 years(thats when the higher return bails you out, also, a lot of the time the decline was unjustified in the first place). That doesn't happen with bonds because it tends to make low returns (which then would take centuries to recover from a huge drawdown). Furthermore stocks have a inflation hedge component in them (which helps that 10-20 year reversion). You only lose it all or close to it when markets get shutdown and don't reopen (Soviet union, Cuba, etc). Those extreme events happen a lot less often than inflation crises, which happen all over the place. Given the difficulty of bonds of recovering from extreme drawdowns (in real terms) and the frequency of inflation crises, I speculate that the average and the median long-term return in government bonds might be negative
With bond ETFs the minimum bet size is very small. Besides it's a very interesting thread, and i've followed your recent post about bonds in the Global macro journal, curious to read counterarguments as well.
Very well... I would suggest to you that you might want to test those statements above which could be classified as hypotheses using the available cases. I have looked at a few and I would have to disagree with at least some of your claims above.
Well, here is a rudimentary test: This is from the book Triumph of the Optimists 1900-2000 The average return is 0.71%. So it appears that bonds have a positive return right? Except this is all the data that is avaliable (at least to those researchers). There are a huge amount of countries where there is no data. But the sucessful countries (like Switzerland, the US, Canada) there is plenty of data. In fact, I dont think there is a stable bond market (as measured in a 100 year period) that is not in that sample, and if there is, its a small number (maybe one or two). But there are a LOT of failed bond markets that are not there. Where is Uganda? Where is Argentina? Greece? The more I look at data, the more I find that success in bond markets is very rare, its the exception not the rule. So, if I had 100 year data of all countries, its very unlikely the return would be positive. Most of the countries not in there (unstable countries), had several currencies, several inflation crises, several debt default crises. Effectively, they repeated the experience of Italy/Germany/Japan/France (which have negative returns), they had one or several huge drawdowns that bond markets need many decades/centuries to recover from. So, it looks to me that its quite unlikely that if had 100y global data, the return would be positive. Its not like sucessful bond markets can remains hidden for long. Sucessful countries become known and turn into 'havens' for capital (even if the country is small), its failure that gets hidden/dissapears
Obviously what Daal writes is disturbing when it comes to traditional portfolio building, Martin, which points do you disagree ?
I'm not necessarily saying investors should avoid bonds, I'm just saying that looking at the data, the risks of large virtually permanent losses are significant. That can be solved through solutions like: 1) A stop loss/exit strategy that avoids a bond market once a specific criteria is met. The criteria is not easy to come up with but it seems quite important. 2)A global bond portfolio that follows a 'market cap weighted' style approach. So it will own more of the bonds that are doing well (in sucessful countries) and less of the ones that are not. Whether the correct weighting is GDP of the country, size of the bond market or something else, is an open question but it needs to reproduce the effect of sticking with success and avoid failure 3) Being good at choosing the right countries
all central banks including the Bundesbank are engines of inflation. can anyone show me an exception? look at the changes in price levels in the the US since the creation of the Fed. in 1913.
there is a book which may prove useful: Triumph of the Optimists: 101 Years of Global Investment Returns plus https://www.researchgate.net/publication/4855012_New_Indices_of_British_equity_prices_1870-1913 Last edited: Jan 29, 2017
Not this again... You know what the solution is? Central banking according to the tenets of proper, full on Sharia law, with its prohibition on riba, will ensure that your central bank will stop being an engine of inflation.
I am not disagreeing with the idea that sovereign bonds can lose money. Obviously, there are risks associated with bonds, as any asset allocation study will be able to tell you. My disagreement is simply with the idea that, in a scenario where you incur large losses in your local ccy govt bond portfolio, somehow you're going to be just fine holding stocks.