Performance Analysis Of Contiguous Swiss Federal Pension Fund Benchmark Indices
Data as of 31 October 2019
Pension fund benchmark indices shown here are based on various benchmark index families calculated and published by Banque Pictet & Cie. But they should not be confused with Pictet’s indices. For the rationale behind Agathos’ contiguous benchmark indices, and a description of how and why they are being calculated, click here.
The performance analysis of benchmark indices serves two main purposes:
- Firstly, such benchmarks give a fair, if very simplified description of the investment environment from which pension funds need to find an optimum blend of risk and reward. An analysis of this environment gives insight into the specific challenges pension funds need to face, which may be comparatively easy during one cycle, but exceedingly difficult during another.
- Secondly, and only in view of the known and extraordinarily high correlation between actual pension fund performance data and such benchmark indices, the monitoring of relevant data with monthly frequency gives ‘live’ insights into the dynamics impacting the asset side of pension fund’s balance sheets. Monthly data, available with little or no delay and put into long-term context, generates an altogether different dimension of understanding, not possible with the annual data available from pension funds themselves, and after months of delay.
Graph-1 (above) shows returns for full calendar years, and current year-to-date values. At the end of October, current year returns are clearly at the upper end of typical positive years. How the remaining two months will impact this, remains to be seen. The striking feature of this graph is the severe erosion in values during 2008. Since then, there have been only two and fairly mild ‘down years’ in these benchmarks. It is debatable whether the correction in 2018 can legitimately be viewed as sufficient to have cleansed financial markets from any perceived, or actual, excess. If at all, such a conclusion could certainly not be drawn on the basis of a very broadly based blend of markets, as general as a global, co-mingled benchmark index. Pension funds now face the conclusion of the elevenths year without a proper bear phase, or general crisis. How Swiss pension funds will manoeuvre the conflict between intoxication from windfall gains and the demands of prudence remains to be seen. So far, each time they were faced with that dilemma, they stumbled over fear of taking decisions at all, and promptly saw capital reserves melt away within weeks. Most, if not all subsequent exhibits seen below will highlight just how benevolent and long-lasting the current investment cycle has been. It should not be taken for granted that it continues.
Below, Table-1 shows index values from 31.12.1999 and at standardised intervals, together with matching annualised rates of return. Already with this crude initial glance it becomes evident that recently the investable universe for Swiss pension funds has been considerably easier than the medium and longer term history. To understand this claim, compare the 12 months rate of with longer time frames, such as five, or even ten years, and the 19 years since 1999. Thus, the obvious challenge for pension funds at this juncture is to ready themselves for a future strategic shift in investment strategy towards capital preservation. When the cycle does turn, it will freeze, just to allow for the unprepared to catch up.
This stealthily accruing scenario is undoubtedly already causing stomach pains across the industry.
Graph-2 (above) depicts performance of the three benchmarks across the most recent 120 months. Considering that all time series have a base value of ‘100’ on 31.12.1999, it is extra-ordinary that they should have converged as much as they have, currently having values comparatively similar to each other, two decades hence.
Table-2 (below) highlights the different route by which each benchmark has arrived at it’s current value. Measured over 120 months, returns have been ‘orthodox’ in the sense that a higher risk benchmark has generated a higher return than a lower risk one. Many market participants falsely view this as a kind of ‘natural financial market law’, which it is not: The expectation of higher returns from higher risk is only valid during times when risk is overstated. When risk is understated, then the unrecognised risk-excess leads to devastating losses. There are two painful reminders of this principle just in the first decade of the current millennium: The implosion of the ‘Dot.Com- Bubble, and the financial crisis caused by supposedly AAA-rated Mortgage Bundles a few years later. The tail end of the mortgage crisis is still contained in some of the metrics calculated in the present analysis.
Graph-3 (above) shows simple benchmark performance for the most recent 10 years (120 months). In contrast to Graph-2 which showed raw data, in this exhibit all data has been set to a base value of ‘100’ for the starting date of the graph. Note how it took more or less three full years before the three benchmarks had properly disentangled from one another again (2013). In the more recent past ( 2013 to 2018), the higher risk-tier benchmark (LPP-C60) has dramatically outpaced the lower risk-tier (LPP-C25).
In Graph-4 (below) the ratio of LPP-C60 to LPP-C25 is plotted. This serves as a simple illustration for the relative performance of equities to bonds. If value of this ratio rises, then equities do better than bonds, and vice versa. Highlighted are the two phases when equities outpaced bonds. The chart illustrates the disadvantages of a static asset allocation.
Graph-5 (above) shows trailing returns over 36 months, annualised. Two points stand out: Firstly, this metric had travelled in deeply negative territory for LPP-C60 for some time during the years 2009 and 2010 (and less negative for LPP-C40). Secondly, returns have been fairly good for all three risk-tiers during the past several years. Again, what is evident in this graph is an unusually extended period of easy returns. All else equal, the odds are increasingly stacked against this to continue, without at least some pretty rough patches around the corner.
In Graph-6 (below) the ex-post risk of all three benchmarks is shown, calculated on the basis of 36 months trails. The graph’s scale shows risk as a negative value. Thus, low risk readings are located at the top, high risk readings at the bottom. It is unusual that from 2012 onward, this metric has remained not only relatively stable, but also at historically low levels. Readers should bear in mind that ex-post measures of investment risk are contrary indicators: Excessively high ex-post risk suggest rock solid support (‘matters can only get better’) Extremely low ex-post risk should be seen as a sign of imminent danger (‘matters can only get worse’). In Graph-6, the latter point is demonstrated quite well around October 2014, when risk readings of all three benchmarks had converged to a rarely seen low level. There is another, very fundamental point to be made from this graph. Unmanaged investment risk (which is what benchmark indices portray and represent) is subject to enormous fluctuations, much more so than adequately managed risk. Thus, adopting a specific benchmark for cloning, and thereby declining to manage risk, inevitably leads into a risk territory, that will probably be much higher, than intended, when a supposedly low risk benchmark was chosen. Passive investing (benchmark cloning) is the antithesis of risk-management.
The beneficial impact of active asset allocation for the purpose of risk management is illustrated here.
Graph-7 (above) is yet another piece of evidence demonstrating how recent and implied concept’s of financial market ‘normality’ should be taken with ‘a grain of salt’. Shown is the net value of medium term normalised returns (as depicted in Graph-5) and medium term normalised risk (as depicted in Graph-6). Put differently, this exhibits plots risk-adjusted rates of return. Risk is not an objective, it is merely a necessary evil. Unless there is a reasonable hope for a correspondingly higher expected return, accepting higher levels of risk is irrational, inefficient and counter-productive. Returns cannot be enforced by higher risk. On the other hand, what matters above all else to pension funds (with nominal future liabilities) is the level of nominal returns, not ‘clean’ returns. But even so, risk efficiency is a qualitative metric. The importance of risk efficiency shows itself in every bear market, when having only the least (necessary) exposure to risk is what avoids large scale long term damage to the asset value, and with it future return potential. On capital losses suffered, no future return can be earned. In recent years (2017 to date) there appears to have been little to no net advantage from being exposed to higher levels of risk.
Table-3 (above) shows normalised annual rates of return and risk and gives the profile of monthly rates of change back to the introduction of mandatory occupational pension schemes in Switzerland. Note how similar returns are over this nearly 35 year long comparison. Annualised rates of return for the three risk tiers fall within a mere 0.1% differential (in favour of the lower risk strategy) whereas risk is more than double for LPP-C60 (5.2% annually) than risk for LPP-C25 (2.1% annually).
Of course, the data refer to benchmarks and thus to unmanaged risk. The point of investment management is to discriminate in favour of ‘cheap’ risk and agains expensive risks which not only reduces overall risk exposure but simultaneously improves returns, if only be preventing meaningful losses along the way.