Exhibit-1 (above): Shown is a plot of yields on ten year benchmark bonds (treasury) for the five markets subject to this monitor. From these yields, return indices are calculated which then serve as the basis of the performance analyses illustrated here. Calculations are for synthetic, zero coupon bonds with a maturity of exactly ten years. German treasury bonds serve as proxy for Euro denominated sovereign debt.
Yields across major world bond markets have been declining, but less so in the US and Great Britain. Swiss government bonds were the first to enter a hitherto unimaginable territory of negative yields (in early 2015), followed by the yield on treasury bonds in Japan and Germany. Currently, three of the five bond markets in the sample have negative yields, pushing the arithmetic average yield of these five markets below 1%.
Declining yields translate into higher bond values and vice versa. Whenever yields will rise again from these very low or even negative yields, the prices of those bonds will be subject to extraordinary losses.
Exhibit-2 (below): The graph shows annualised rates of return for several time frames, comparing the composite index of five major bond markets with LPP-C40, a popular multi-asset benchmark for Swiss pension funds. For the full history since occupational pension funds became mandatory (31.12.1984) , bonds have performed better than this benchmark. The same applies to the time since the end of 1999. However, in the more recent past, bond returns have lagged equities. A comparison between the bond composite and LPP-C40 benchmark index was negative across ten, five, three years, and during the last 12 months.
Exhibit-3 (above): The table focusses on returns over ten years and less. In the upper part, values of return indices are shown, while the lower part displays annualised rates of returns for matching time frames since these dates. For greater transparency, the best returning market in each period is shown in green, the worst performing in red.
Over ten and five years, British bonds show the highest return. Across anything from 36 months and less, US Treasuries generated the highest returns in the sample. January 2020 was a strong month for bonds, across the board.
Exhibit-4 (below): The graph emphasises returns over the most recent 12 months, a period clearly dominated by good returns from US treasuries while returns form Japanese bonds were very low. British bonds, the second best bond market, generated only slightly more than half the return from US Treasury bonds.
Exhibit-5 (above): Returning to the bond composite, the graphs above plot straight performance in the lower portion of the exhibit, and the relative performance to the multi-asset benchmark LPP-C40. As an asset class, for several months now, bonds have been a drag on performance, if not in absolute, then in relative terms. Currently, the relative plot stands very near its long term low reached late 2018.
Bonds play a very important role in the pursuit of stable long term returns of diversified portfolios. But this is entirely reliant on comparatively short periods of time, when bonds generate meaningful performance over and above equities, not the other way around. Bonds should be a tactical investment, not a strategic core.
Exhibit-6 (below): This graphs covers the same time frame as the preceding plot of the composite index but this time individual bond markets are shown. Note the massive performance gap between British and German bonds (the strongest two bond markets) and the Japanese bond market, the laggard. Relative performance of individual bond markets will be emphasised further down, in Exhibit-9.
Exhibit-7 (above): The table shows values of the all return indices, and their performance differential to the equal-weight composite in the upper portion. Over ten years, British Treasury bonds performed considerably better than the other markets in this sample, Japanese Treasury bonds had the poorest return in this time frame.
In the lower part of the table, the elements of a regression analysis are shown: Alpha (intercept with y-axis) is shown in green when positive, red when negative. Beta (slope of regression line) is shown in green when below 1, and in red when above. The regressions cover 120 monthly rates of change.
Remarkable are the low values for r-squared, not even reaching 50%, national bond markets are far less correlated to one another than national equity markets are.
Exhibit-8 (below). Scatters show bond markets (and the ‘M5’ composite) in terms of normalised return (vertical) and normalised risk (horizontal). To the left, latest 36 months are shown, to the right values for the latest 60 months.
The dashed diagonal lines indicate equality between return and risk.
Clearly, in the more recent past, bond markets have been more rewarding than at early in the most recent five year history. All markets have moved to the left (reduced risk) and up (higher returns). That said: Japanese, German, and Swiss had a diluting effect on efficiency.
Exhibit-9 (above): This display emphasises differences in performance by plotting values relative to the composite index. It covers a shorter period of time than most exhibits shown here, only three years. The two markets with the highest coupons (still positive), British and US Treasuries were the only ones to have generated positive relative returns. The strength of the US bond market only started relatively recently, in the third quarter of 2018.
Exhibit-10 (below): The table examines returns over three years in the upper part. Except for US Treasuries, return indices are mildly below their peaks. Only US Treasuries have produced risk-adjusted returns in positive territory. Swiss Treasuries have the highest ex-post risk, followed by their British peers. While risk on Japanese bonds has been very modest, the returns are too low to generate positive adjusted returns.
The monthly changes in all five bond markets are heavily distorted horizontally.
Exhibit-11 (above): While the chart covers 10 years of history, it displays trailing returns calculated over 36 months each. With considerable fluctuations, this metric has been trending downward for all five bond markets. Only recently, lost of 2018) have three-year total returns from bonds dipped into negative territory. Interestingly, US Treasuries, the market with the highest yield, was the first to do so, prompting a rally thereafter. As an asset class, bonds seem to be gaining momentum recently.
Exhibit-12 (below): Illustration 12 shows ex-post risk (‚observed risk’) calculated over 36 months long trails. It is quite remarkable that typically, the US and Japanese bond markets differ so widely by this measure, with Japanese bonds containing considerably less historic risk than their American peers, in spite of the much lower yield. Care must be applied in the interpretation of this or any other ex-post metric. Risk, just like return, is highly variabel and forward looking risk may be substantially different from the ex-post analysis. Generally speaking, low levels of ex-post risk are usually a warning sign, as risk is bound to resurface sooner or later.
Exhibit-13 (above): To conclude, this chart shows annualised rate of returns for the 20 calendar years since 1999. The advantage of US Treasuries looks less pronounced, with 20 year returns being almost identical to British treasury bonds and not much higher than German treasuries. Japan and Switzerland have performed poorest within this group, largely from their much lower yields.