US (S&P 500) Stock Market – December 2009 Update

To try and squeeze some more performance out of a retirement investing strategy that is heavily focused on asset allocation I am using a cyclically adjusted Price / Earnings (PE) ratio for the S&P 500 to attempt to value the US (specifically the S&P 500) Stock Market. The method used is that developed by Yale Professor Robert Shiller. My numbers will always appear slightly different to Professor Shiller as I also use the Earnings forecasts from Standard & Poors to complete the data set up to the month of interest. I will call it the Shiller PE10 and it is the ratio of Real (ie after inflation) S&P 500 Monthly Prices to 10 Year Real (ie after inflation) Average Earnings. I will use this data set as a proxy for my UK Equities as well as my International Equities (mostly Japan, US and EU equities). Ideally I would create a PE10 for each country however good data is impossible to find and I am working on the principle that these stock markets are now heavily correlated with globalisation.

To give an example of this I have just read a very good book called Wall Street Revalued by Andrew Smithers. He presents data that shows the following stock market correlations for the period 1999 to 2008:
US to France, 0.92
US to Germany, 0.93
US to Japan, 0.84
US to UK, 0.93

I will assume that stocks always return to their long run PE10 average and so reduce my exposure when stocks are overvalued and increase my exposure when they are undervalued. For my US and International Equities I will use the long run average of the Shiller PE10 to equate to when I hold 21% UK Equities and 15% International Equities respectively. For my retirement investing strategy on a linear scale I will target 30% less asset allocation when the Shiller PE10 average is Shiller PE10 average + 10 and will have 30% more asset allocation when the Shiller PE10 average is PE10 average - 10.

All figures are taken from historic data provided by Professor Shiller. Current stock market data is taken from Standard & Poors and inflation from the Bureau of Labor Statistics. The December market price is the 30 December S&P 500 stock market close.

Chart 1 plots the Shiller PE10. Key points this month are:
Shiller PE10 = 20.6 which is up from 19.8 last month. My UK Equities target asset allocation is now 18.3%. Additionally my International Equities target asset allocation is now 13.1%.
Shiller PE10 Average = 16.4
Shiller PE10 20 Percentile = 11.0
Shiller PE10 80 Percentile = 20.6. This means the S&P 500 is currently sitting right on the 80 Percentile.
Shiller PE10 Correlation with Real (ie after inflation) S&P 500 Price = 0.78

Chart 2 plots further reinforces why I am using this method. While the R^2 is low there appears to be a trend suggesting that the return in the following year is dependent on the Shiller PE10 value.

Chart 3 plots Real (after inflation) Earnings and Real Dividends for the S&P 500. Real Dividends are still falling however they are still above their long term trend. Real Earnings have a roller coaster ride continually, particularly since about 1990. If the Standard and Poors forecast earnings are to be believed however we are now back above the long term earnings trend.

US Inflation – November 2009 Update

The above chart shows the Consumer Price Index (CPI-U) up to November 2009 courtesy of the Bureau of Labor Statistics. I have taken the liberty of dividing the chart into two sections. The first red section runs from 1871 to 1932 and the second blue section runs from 1933 to present day. I chose this break point as during 1933 the US officially ended their link to the gold standard. I think this chart demonstrates a point that government will always choose to inflate debt away at the expense of savers if given the chance. They could not do this under the gold standard. To demonstrate this average inflation rates have been:
1871 to 1932, 0.5% with deflation being a regular occurrence.
1933 to Present, 3.7%

Australian Stock Market – December 2009 Update

To try and squeeze some more performance out of a retirement investing strategy that is heavily focused on asset allocation I am using a cyclically adjusted PE ratio for the ASX 200 to attempt to value the Australian Stock Market. The method used is based on that developed by Yale Professor Robert Shiller. I will call it the ASX 200 PE10 and it is the ratio of Real (ie after inflation) Monthly Prices and the 10 Year Real (ie after inflation) Average Earnings. For my Australian Equities I will use a nominal ASX 200 PE10 value of 16 to equate to when I hold 21% Australian Equities. On a linear scale I will target 30% less stocks when the ASX 200 PE10 average is ASX 200 PE10 average + 10 = 26 and will own 30% more stocks when the ASX 200 PE10 average is PE10 average -10 = 6.

All figures are taken from official data from the Reserve Bank of Australia except December which is estimated by taking the price from the 28 December Market close. Additionally the December Dividends and Earnings are assumed to be the same as the November numbers.

Chart 1 plots the ASX 200 PE10. Key points this month are:
ASX 200 PE10 = 18.7 which is up from 18.5 last month. My target Australian Equities target is now 19.3%.
ASX 200 PE10 Average = 22.9
ASX 200 PE10 20 Percentile = 17.3
ASX 200 PE10 80 Percentile = 27.7
ASX 200 PE10 Correlation with Real ASX 200 Price = 0.82

Chart 2 plots further reinforces why I am using this method. While the R^2 is low there appears to be a trend suggesting that the return in the following year is dependent on the ASX 200 PE10 value.

Chart 3 plots Real (after inflation) Earnings and Real Dividends. Dividends appear just about on trend however Earnings are still struggling and heading very much downwards. If this continues and Prices hold the ASX 200 PE10 will surely start to rise.

UK Inflation – November 2009 Update

The Office for National Statistics (ONS) reports the November 2009 UK Consumer Price Index (CPI) as 1.9% and the UK Retail Price Index (RPI) as 0.3%.

On the surface this sounds ok as the Bank of England has the following remit:
“The Bank’s monetary policy objective is to deliver price stability – low inflation – and, subject to that, to support the Government’s economic objectives including those for growth and employment. Price stability is defined by the Government’s inflation target of 2%.” The inflation measure they use is the CPI and so why would they raise interest rates?

I however don’t like what I see when I look at the raw data and think inflation could quickly get out of hand given the very low interest rates and quantitative easing that is currently occurring.
The first chart is tracking the CHAW Index which is the RPI including All Items. I focus on the RPI as my National Savings and Investments Index Linked Savings Certificates use the RPI to index from. This saw a big dip when the Bank of England dropped interest rates to historic lows however the chart shows that all the dip did was compensate for the big kick upwards that was seen from 2007. The current level of the Index is just about on the trendline suggesting we are back to the average annual increase since 1987 which is around3.5%.

The second chart is again based on the CHAW Index. This chart shows annual figures based on the previous 3, 6 and 12 month’s worth of data. As of November the 12 month figure is 0.3% (as published by the ONS) however disturbingly the 6 month figure is 3.6% and the 3 month figure is 4.1%. It will be interesting to see the official January 2010 numbers which is 12 months from the low in the Index at January 2009 and whether the Bank of England does anything about it. My simple projections suggest this could be up to 3.7%. I’ll keep buying Index Linked Savings Certificates whenever possible if this is going to continue.

Tax Efficient Investments and Tax Wrappers

You would think that governments throughout the world would be trying to encourage people to save for their own retirement and hence remove future burdens from the state. Unfortunately in the United Kingdom I don’t see the current government providing much support or encouragement here. Instead I see a lot of complexity which means unless you know exactly what you are doing you will over your investment life either:
- pay higher taxes than are necessary thereby ending up with a smaller investment pot in the future, or
- pay the same (effectively tax deferral) or lower taxes within tax wrappers but lose a lot or all of this benefit by paying higher fees / charges than needed or even in extreme cases charges on charges with in some instances the added negative of having certain restrictions on what you can and can’t do as defined by the government. Just to give one extreme example, I could even see that if you are only deferring tax within a wrapper but paying higher charges then you are actually worse off than not having a tax wrapper at all and just paying the full tax.

I guess it’s no different to anything else in life. Buyer beware and always carefully do your own research.

Minimising the tax I pay (along with the fees /charges I pay) is one of the cornerstones to my retirement strategy. I can easily show this by thinking of what percentage of my annual returns will come through dividends / interest and the affect that paying high rates of tax will have on my final retirement fund through losing on the compound interest effect. Let me give an example. Average Joe and Mr UK Average have £10,000 in share based investments that provide a dividend yield of 4.5% (average dividend yield of the S&P 500 since 1871 using the Shiller dataset). Both are higher rate 40% tax payers. Now Mr UK Average who doesn’t think about minimising taxation holds these shares outside of any tax efficient wrappers and so pays 32.5% (however in most instances this is really only 25% after tax credits on dividends are calculated). Average Joe holds half of his share based investments in tax wrappers meaning his effective tax rate is only 12.5%. I’m going to ignore capital gains to make the demonstration simple. So where are they after 20 years:

Average Joe has 11% more assets (and the more years in the calculation the better it gets for average Joe) than Mr UK Average meaning he has the potential for 11% more income in retirement or could hold a less risky investment portfolio for the same lifestyle. Taxes on investments matter.

I use one tax efficient investment type which are National Savings and Investments Index Linked Savings Certificates. Additionally I use 2 tax wrappers which are a Pension which is a Defined Contribution Scheme and a Stocks and Shares Individual Savings Account (Stocks and Shares ISA).

Just a little about each of these as I will cover them in more detail later:
- NS&I Index Linked Savings Certificates today offer a return of the Retail Prices Index (RPI) + 1% however the real benefit, particularly for high rate tax payers is that they are tax free.
- Pensions are extremely complex things and you would think that any government would be trying to simplify them so they could be understood and make them beneficial for all tax payers to save for their future. Unfortunately this doesn’t seem to be happening. I use them as for me personally they provide huge benefits. I can however think of a number of situations where they would provide little to no benefit and in some instances could even make a person worse off. What I am always nervous about however is that with a Pension you lock your money up for a lot of years and with past history as a guide governments will most likely change the rules. I therefore don’t put all my investments here.
- Stocks & Shares ISA’s seem to be no brainer for anyone wanting to invest in the stock markets. They form a pivotal part of my investment strategy. If you look around you can find Stocks and Shares ISA’s that don’t charge you to use the tax wrapper. You still pay to buy the investments inside and these investments might also see charges however this is no different to outside the wrapper. For higher rate tax payers they provide tax benefits from both dividends and capital gains. While for low rate tax payers today, they offer only really capital gains benefits. One thought however is that you never know when you might become a higher tax rate payer or given the disastrous state of the public finances of the UK you never know when as a low rate tax payer you will be taxed on your dividends.

Building My Low Charge Investment Portfolio – Part 3 of 3

Portfolio construction starting point as previously described in Part 2:

22% Bonds (Cash, Index Linked Gilts, Index Linked Savings Certificates)
21% Australian Equities (ASX 300)
21% United Kingdom Equities (FTSE All Share)
15% International Equities (40% US, 40% EU, 20% Japan)
5% Emerging Markets Equities (MSCI Emerging Market TRN Index)
10% Property (Listed Euro Property and UK Commercial)
5% Commodities (Gold)

The final element associated with the construction of my portfolio is an attempt to assess whether the stock market is overvalued or undervalued. I am doing this as the stock market can be a highly volatile / high risk place and if I can take a calculated risk here I might be able to squeeze some more performance out of my portfolio. I will try and be underweight equities when the market appears overvalued and overweight equities when the market appears undervalued. This could be likened to timing the market which history suggests is difficult/impossible meaning I may also end up under performing compared with the traditional buy, hold and rebalance periodically strategy. Only time will tell.

The method I am going to use is that developed by Yale Professor Robert Shiller. He uses a very simple method where he looks at the monthly real (after inflation) market price of the S&P 500 and divides it by the average of the previous 10 years real earnings to get a long term historic real price earnings ratio (Shiller PE10). The real price and Shiller PE10 look to have a relatively high correlation of 0.78. A chart showing both the Shiller PE10 ratio versus the real S&P 500 can be seen below. The December 09 entry is the market close on the 04 December. My current Shiller PE10 estimate will always appear slightly different to Shiller as I use the Standard & Poors website to also enter forecast earnings estimates up to the month of interest. Since 1881 the average Shiller PE10 value has been 16.35.


Another way I have looked at this is to plot a scatter diagram of the Shiller PE10 versus the return made in 12 months time shown below. This I have calculated as the ((Price in 12 months – Price)/Price)+(Dividend in 12 months)/Price). While the R^2 value is quite low a trend line suggests there may be something to be had.


While I will use the average Shiller PE10 value I won’t go silly. I will try to squeeze some performance by on a linear scale will owning 30% less stocks when the Shiller PE10 average is Shiller PE10 Average + 10 = 26.35 and will own 30% more stocks when the Shiller PE10 average is PE10 average -10 = 6.35. I will use this methodology as a proxy for all my International Equities and United Kingdom Equities as Hale suggests a high correlation between UK Equities and International Equities and I have struggled to find good historic data for UK, European and Japanese equities. Today’s Shiller PE10 is 20.2 meaning that instead of holding 21% UK Equities today my target is 18.6%. Similarly instead of 15% International Equities today my target is 13.3%.

For my Australian Equities I have been able to find some Inflation and ASX 200 data from the Reserve Bank of Australia. Unfortunately all the data needed only starts from 1982 which is not a long time ago. Even so I have calculated and show what I will call the ASX 200 PE10 chart versus the Real ASX 200 below. I have estimated average ASX 200 PE10 to November 2009 to be 18.5. The correlation between the Real ASX 200 and the ASX 200 PE10 appears high at 0.82. For my Australian Equities I will use a nominal ASX 200 PE10 value of 16 to equate to when I hold 21% Australian Equities. Otherwise I will use the same assumptions as for the UK and International Equities. Therefore with an ASX 200 PE10 of 18.5 today my target is 19.3%.

When I am underweight I intend to hold the extra in bonds/cash. For my emerging markets equities because it is a small amount I will not vary the weighting but always target 5%.
So that’s how I’ve arrived at my target asset allocations. Today this means a Desired low Charge Portfolio of:

28.8% Bonds (Cash, Index Linked Gilts, Index Linked Savings Certificates)
19.3% Australian Equities (ASX 300)
18.6% United Kingdom Equities (FTSE All Share)
13.3% International Equities (40% US, 40% EU, 20% Japan)
5% Emerging Markets Equities (MSCI Emerging Market TRN Index)
10% Property (Listed Euro Property and UK Commercial)
5% Commodities (Gold)

My Current Low Charge Portfolio – December 2009

Another month passes.
Buying: Property, Index Linked Gilts, UK Equities, International Equity
Selling: Nothing this month
Dividends: Nothing this month
Current UK Retail Prices Index: -0.78%
Current Annual Charges: 0.60%
Current Expected Annual Return after Inflation: 4.26%
Current Return Year To Date: 22.9%
How close am I to retirement: 40.8%

Building My Low Charge Investment Portfolio – Part 2 of 3

Portfolio construction starting point as previously described in Part 1

28% Bonds
72% Equities

Continuing to learn from Hale and Bernstein I then chose to add Property as a portfolio diversifier as it can be uncorrelated to Equities and Bonds. The more uncorrelated the assets theoretically the more likelihood that some assets will be low in price (and hence required to be bought by the author) while other assets are high in price (and hence required to be sold). To give 2 examples, Bernstein presents data from 1973 to 1998 suggesting a Property – (REIT - National Association of Real Estate Investment Trusts) to US Equities correlation of 0.56 while Hale presents data suggesting a UK Property to UK Equities correlation of 0.3. The property types I have chosen to buy into are predominantly Listed European Property Securities excluding the United Kingdom and a fund buying United Kingdom commercial property. I chose to allocate 10% to this sector. Portfolio is now:

23% Bonds
67% Equities
10% Property (Listed Euro Property and UK Commercial)

Next I chose to add some commodities exposure. Initially I went for exposure via a low cost ETC that gave exposure to energy, precious metals, industrial metals, livestock and agriculture via Futures contracts. In hindsight I think that what I did was a mistake and the reason you should always do your own research as I lost quite a lot of money here. Before anybody buys direct exposure to commodities I suggest you read about Contango / Backwardation , Counter-party risk. Additionally many ETC / ETF’s allow you to leverage your returns. Be sure to understand the affects of this also as they are typically rebalanced daily meaning performance will differ from what you expect. So what did I buy? In the end I settled on a simple vanilla ETF that buys physical gold and so tracks the gold spot price minus fees. Why gold? Using the Shiller dataset for the S&P 500 and the Austin Rare Coins dataset for Gold for the years 1968 to present day a relatively low correlation of 0.5 is achieved. What does 0.5 look like? Well rather than just show actual prices in history I present a chart showing real prices (ie after inflation) for both assets.


Additionally, it’s also worth looking at a chart showing the ratio of the S&P 500 to Gold (excluding inflation).


As always make up your own mind but I’ve settled for 5% to be allocated to Gold. As commodities can be highly volatile I’ve taken my Gold allocation from equities. Portfolio is now:
23% Bonds
62% Equities
10% Property (Listed Euro Property and UK Commercial)
5% Commodities (Gold)
Now to how I’ve allocated my equities. Hale suggests that traditional developed world equities have very high correlations of about 0.9 and so do not significantly diversify the portfolio. I guess this is because of globalisation. Emerging Markets on the other hand seem to have a much lower correlation of around 0.7. I’ve still chosen to spread widely through the world to remove country risk as much as possible and also to pick up some swings from exchange rates. I’ve settled on 21% Australian Equities, 21% UK Equities, 15% International Equities and 5% Emerging Market Equities. My Australian Equities are generally ASX 300 low cost trackers. My UK Equities are generally FTSE All Share low cost trackers. My International Equities are targeted to be 40% US, 40% Europe and 20% Japan. My 5% Emerging Markets simply track the MSCI Emerging Market TRN Index. Portfolio is now:
22% Bonds
21% Australian Equities (ASX 300)
21% United Kingdom Equities (FTSE All Share)
15% International Equities (40% US, 40% EU, 20% Japan)
5% Emerging Markets Equities (MSCI Emerging Market TRN Index)
10% Property (Listed Euro Property and UK Commercial)
5% Commodities (Gold)
Finally to my bond allocation. Within my bond allocation the first thing I do is hold 6 months worth of salary in cash. Just in case... Next within low tax wrappers I’m holding predominantly holding Index linked Gilts as I’m always concerned about returns after inflation. The final variation from a traditional bond allocation is that outside of the tax wrappers where instead of bonds I am holding what I consider to be a great little product from National Savings and Investments which are Index linked Savings Certificates. Portfolio is now:
22% Bonds (Cash, Index Linked Gilts, Index Linked Savings Certificates)
21% Australian Equities (ASX 300)
21% United Kingdom Equities (FTSE All Share)
15% International Equities (40% US, 40% EU, 20% Japan)
5% Emerging Markets Equities (MSCI Emerging Market TRN Index)
10% Property (Listed Euro Property and UK Commercial)
5% Commodities (Gold)

So that’s my portfolio based on what I think are basically traditional asset allocation strategies. The final portfolio weightings however are where I try and introduce allocations that vary based on what I think is a sensible method for determining asset values at any point in time. This I’ll cover in Part 3.

Building My Low Charge Investment Portfolio – Part 1 of 3

Note added 11 January 2009. My calculations below generally use arithmetic means. It may be more appropriate to use compound annual growth rates. See here.

Before I start let me emphasise that my strategy is a long term one. I am not a share trader. In fact, I tried it very briefly and was hopeless. For me the duration of my investment strategy is the rest of my life as I’m using it for retirement planning. Retirement in my definition is that work becomes optional and so I’m trying to get there as quickly as possible. However I guess it might also work for other long term goals – university fees for a newly born child, planning to give a newly born child a head start on their 18th birthday etc.

Now the first thing I did was some reading. I would like to recommend two very good books – “Smarter Investing : Simpler Decisions for Better Results” by Tim Hale and “The Intelligent Asset Allocator” by William Bernstein. If you are UK based I’d start with Hale and if US based then Bernstein. I can’t highlight enough how beneficial these books were in my retirement strategy.

The first two building blocks I chose were equities and bonds.

What returns can I expect from equities and bonds? Firstly I will always try and talk in after inflation terms (I’ll typically use the word Real) as we all want to maintain our standard of leaving. Please don’t underestimate the damage of inflation to your wealth. To demonstrate how damaging inflation is I will use a very good data set that is available from Yale Professor Robert Shiller’s website1 that lists US CPI data. This demonstrates that to have the equivalent of $12.46 in 1871 today would require you to have in your hand $216.24. This has made me always consider inflation in everything I do.

So to equity returns. Hale suggests that for the UK from the period 1900 to 2004 real (ie after inflation) average equity returns were 7.3%. Bernstein suggests that for US 1926 to 1998 equity returns averaged 11.22% however again using the Shiller1 data set as a basis suggests average inflation over that period was 3.2% so a real return of 8%. Using the Shiller dataset from 1871 to present day for US equities suggests that the real average capital gain was 3.5% and the real average dividend was 4.5% for an average real return of 8%. So as close as I need it to be for what I am doing.

Now to bond returns. Hale suggests that for the UK from the period 1900 to 2004 real average bond returns were 2.3%. Bernstein suggests that for US 1926 to 1998 5 year treasury returns averaged 5.31% for a real return of 2.11%. Again as close as I need it to be.
Finally combining equity and bonds as my basic building blocks and having assessed my attitude to risk I came to a 72% equity and 28% bond weighting. In multiple locations I have heard a ‘rule of thumb’ that you should own 100 minus your age in equities. I’m more aggressive than that at this stage of my life.

In projecting retirement dates I’m assuming real average equity returns of 7.3% and for the majority of my “bonds” real average returns of 1%. The 1% I’ll explain in the future when I look at some strategies I’m using to minimise the tax I pay. Assuming I stayed at this allocation and achieved averages I could assume an instantaneous real average return of 72% x 7.3% + 28% x 1% = 5.5% - 0.6% (my current average fees) = 4.9% per annum before taxation. Now clearly I won’t get this consistently as I’ll reduce equity weightings as I get older and nothing works in averages but it demonstrates an order of magnitude. Worst case we have a big crash in equities and/or bonds and I get a return much worse than this. Hale shows the effects and risks of this quite nicely.

Now by rebalancing regularly to take advantage of the fact that over the long term equities and bonds are not perfectly correlated I’ll end up buying whatever is lowest in value and selling whatever is highest to hopefully give a little more of a kick.
I could have probably left it at that and hopefully lived happily ever after however as I’ll show as I blog I’ve gone a lot further than that.

To be continued in Part 2 here