December 14th, 2016
Hugh Hodges from Harpers Bernays shares his thoughts for trustees and many NFP organisations who currently have Bonds in their portfolios who are currently asking the same questions.
A significant number of NFP investment portfolios hold a minor, or occasionally a majority, portion of their portfolio in fixed interest securities, either directly or normally via bond funds. In most cases this is to limit the likely occurrences of a fall in the total portfolio value over a 12 month period, or otherwise limit the volatility of portfolio as a whole. In some cases this is due to rules imposed on them, in others it is a choice for reasons such as limiting volatility of portfolios to which donors contribute, in part to prevent the impression that some of the donated funds have been “wasted” due to “inappropriate” investment of the funds.
If it is being done due to the potential of a significant drawdown from the portfolio it is likely that the portfolio should be restructured to keep any amount likely to be drawn down in a separate part with a different asset structure, however that is another issue.
In almost all circumstances in the past 25 to 35 years, investing a part of a long term portfolio in bonds has lowered the overall income that could be generated by the portfolio, and has reduced its ability to grow over time and thus keep up with inflation, but it has contributed to lower volatility of the portfolio as a whole and done so in a way that produces a higher return than if that portion had been held only in cash.
Most portfolios have had some sort of modelling done in the past to determine the appropriate mix of higher income, but more volatile, “growth” assets such as Australian shares, and “defensive” (but in recent times lower income producing) cash and bonds. In most cases the modelling was done using a government bond index as a proxy for fixed interest, cash using the cash rate average returns and Australian shares with an ASX share index. Sometimes other asset classes such as international shares and bonds or property were also included in this analysis.
In almost all cases the returns and other data used for that modelling were historical returns – ie backwards looking data. In almost all circumstances that approach is entirely appropriate, however the time that it can be inappropriate to use is when a market is at a very long term, or secular, turning point. As argued below this may currently be the case with fixed interest assets in which case the use of historical data in formulating asset allocation structures may produce an entirely misleading outcome.
I strongly believe that now is an appropriate time to revisit the asset allocation of all portfolios that contain large allocations to bonds, especially NFP portfolios that often have bonds there for slightly different reasons than traditional tax paying portfolios. This is because the assumptions and inputs into the models used in the past decade, or more, especially in relation to likely returns from cash and, in particular, bonds probably have changed since then. It may also be the case that the return assumptions may no longer be appropriate for these portfolios going forward due to the methodology used by the advisors at the time. There are a number of different ways of forming the risk and return assumptions based on client directions / needs but almost without exception they are backwards looking on the basis that that is the only way that objective numbers can be created. This is most especially the case if a secular market trend (such as that which has existed for bonds for more than a quarter of a century) has a high probability of being reversed in the near term.
To illustrate why I advocate a slightly different (and possibly less scientific but I would argue more relevant) approach at this time it is important to understand where we are in the very long term inflation / interest rate cycle which was extended due to the impact of the GFC. The chart below shows the peak in interest rates in the 1980’s, and the subsequent fall since then which was accompanied by a fall in inflation (a global phenomenon). It also shows the renewed fall after the GFC in 2008/09 to historically low rates after appearing to bottom in the mid 2000’s.
Given interest rates in many offshore countries are now around zero (and below in parts of Europe) and the global economy is starting to emerge from the ravages of the GFC, it is a reasonable assumption that both the current low levels of inflation and interest rates cannot continue for much longer (at least in terms of the time frames relevant for looking at asset allocation decisions).
In addition, the election of Donald Trump to the Presidency of the USA (happening as I write this) has the potential to be a substantial game changer for inflation and interest rates in the US, and then across developed economies, if his apparent policies of substantial unfunded fiscal expansion are carried out. This would be very pro-growth and pro-inflation which would push global interest rates back up. At almost any other time in an economic cycle it would be seen as irresponsible and negative by markets, but in the current post GFC world it could be the equivalent of the “New Deal” in the 1930’s which helped stimulate the US economy out of the great depression. The risk is the apparent policies in regard to international trade have the potential to have precisely the opposite effect, so at this stage all that can be said about the changes in the USA is that they have the potential to be very negative for bonds at a time when other forces are already stacking up to be negative for bonds (i.e. potential for interest rates to rise).
The importance of looking at this is that as interest rates fall, the return on a bond portfolio is increased via the capital value of the bonds increasing (at least on a “mark to market” basis, which is how the bond funds record the value and the performance data used in analysis of this type). However, if they rise the opposite happens.
In the past 35 years we have seen the greatest fall in interest rates that we are likely to see during anyone’s current lifetime, and possibly the greatest that has ever been seen in history. Furthermore, nominal interest rates are now lower than they have ever been, looking at records going back to the 16th century. Real interest rates are in many developed countries, negative. So it is a reasonable assumption that, from here, the most likely direction of long term interest rates (and inflation) is up – the only question is by how much and how soon will it start.
The relationship between bond rates and inflation is not direct. Rates try to predict where inflation will be, although in the 1970’s when there was a huge spike in inflation, the interest rates were regulated by the government and did not rise anywhere near where they would in an open market like we have today. The following chart shows a much longer period and shows how unusual inflation is in the current period compared to the rest of our history.
The importance of this, both for the modelling of asset allocation and for potential returns, is that this fall in interest rates has added to the returns on bonds for all periods going back for 35 years, but in the next decade, or perhaps more, not only are the extra returns it has produced not going to be there, capital losses will be incurred which will further exacerbate negative absolute returns as inflation and interest rates rise.
Any modelling based on historical figures will always potentially mislead as to the returns available in the future. However, this will be most especially the case where historical secular trends lead you to make decisions where judgement and experience suggest that long term trends are at a “flex” point for a particular asset class where slavish adherence to the historical data might lead to less than optimal outcomes in the foreseeable future.
There will still be good diversifying benefits from cash and bond type assets if the objective is to reduce total portfolio volatility, but the asset class to use to diversify is likely to be cash and term deposits rather than bonds, given highly probable near term developments.
To get a sense of this and the potential impact on relative returns between cash and bonds in the future, I have prepared the following table of cash (RBA cash rate), bond and TD returns over different periods up to the end of October 2016.
From the table it is easy to see where, 5 to 15 years ago, a figure of 6% could be put into a model for bonds and a figure of 3% to 4% for cash. However that is history that is unlikely to be repeated in the near future, or possibly for a much longer period. Given that cash and bonds provide roughly the same diversification benefits, and bonds have produced a return of between 2% and 3% per annum more than cash over the past 10 to 15 years, it made sense to have bonds as the major diversifier……… until now.
However while in the very long term bonds should produce a slightly higher return, on average, than cash, due to the lower liquidity and slightly greater risks involved in holding them, that premium should be measured in fractions of a percent per annum, not 2% to 3% pa. The rest of the difference is due to the impact of continuously lower interest rates over the period on the returns for bonds.
However, from here, the direction of interest rates is likely to be up for a considerable time, although not back to anywhere like what was seen in the 1980’s, at least in the timeframes we are considering. Therefore, the excess returns that bonds gave over cash will be reversed as rates go back up.
If the excess returns over the past 5 years (say from just after the worst of the GFC) is 3.3%pa it is possible that 3% pa of that could be reversed over a similar period. That would be a 6% turnaround: instead of a 3% tailwind the bonds could produce a 3% headwind compared to cash returns producing near zero, or perhaps negative total returns over an extended time period. That would only put 10 year bonds back to a bit over 4% from their current 2.2% (although they were 1.8% only a couple of months ago and as I write on the day after the election they are up to 2.5%, with the commensurate implied capital losses which go with that) and cash rates are back to about 4.25%. In that case bonds would produce about 3% pa less than cash over that time, although it may well occur over a shorter time, as markets tend to anticipate and react quickly.
On that note, it is instructive to see the difference between bond and cash returns in the past 12 months even though bond rates have moved up a bit lately.
In addition bank term deposits can and have produced a return significantly over the cash rate in recent years as shown in the table above. This is because of the new regulations affecting bank capitalisation and funding sources where they are forced to pay a premium for “safe” non- institutional local deposits.
Looking forwards, rather than backwards as most quantitative models do, it would make a lot of sense to hold term deposits as a diversifier from the volatility of equities instead of holding bonds for that role for the foreseeable future.
The following chart shows the returns from bonds (total return including capital gains and losses) and cash (the interest only, as there is no capital movement) on a year by year basis (to the end of October each year).
The table above shows that bonds have outperformed cash substantially over the 25 years shown, as they have since interest rates peaked in the early 1980’s. However, as it shows the data on a year by year basis you can see during the period there were a couple of 12 month periods where cash substantially outperformed bonds and a period during the first half of the 2000’s where the result was line ball. Those are the short periods of rising interest rates during that long secular period of falling interest rates. In particular, the 1994 year and the 1999 year, where bonds produced a negative total return, correspond with the two periods of rising interest rates shown on the first chart.
While the purpose of this paper is primarily to argue whether bonds or cash /TD’s are the best asset class to use as a diversifier against the volatility of equities in the portfolio, it is worth looking at how Australian equities, which in the most part they are in portfolios to diversify the returns of, performed over the same periods.
Over the 25 years shown in the chart above there were 6 individual years where the share index, including dividends, produced a negative return for the year. In two of them, 1992 and 1994 where shares fell by 12.9% and 1% respectively, cash substantially outperformed bonds and would have been a better diversifier than bonds in terms of impact on total portfolio return. In 2002, where shares fell by 2.2%, there would have been no material difference.
In the GFC year, 2008, holding bonds made a major difference. Shares fell by 37.8%, cash produced 7% and bonds produced 13% as interest rates collapsed. On a 55% equities/45% debt portfolio, holding that debt all in bonds would have shown a total portfolio return of -14.9% but -17.6% if the debt was all in cash. In the other two years, 2011 and 2015, where equities fell by 3.7% and 0.7% respectively, bonds substantially outperformed cash, but by holding either bonds or cash the total portfolio still would have produced a positive return.
Director – Philanthropy
Harper Bernays Limited
10 November 2016
DISCLAIMER: Harper Bernays Limited has made every effort to ensure that the information contained in this presentation is accurate. This document is not intended to provide a comprehensive description of the Harper Bernays’ investment, its investment opportunities or its investment strategies.
Harper Bernays Limited has not taken into account the investment objectives, financial situation or particular needs of any particular client. Before making an investment decision on the basis of this information, you should consider whether the information is appropriate in light of your particular investment objectives, financial situation or particular needs. You may wish to consult external advisers to assist you with your decision.
Sign up for a free edition of our weekly newsletter to have sector news, views and resources delivered straight to your inbox.