Friday 13 May 2011

The ASX 200 cyclically adjusted PE ratio (ASX 200 CAPE or ASX200 PE10) – May 2011 Update

Since simultaneously getting back into the blogging world and sorting my Retirement Investing Today strategy I have already looked at the valuation of the UK Equities market (proxy FTSE100 CAPE) where I personally hold 16.8% of my net worth.  Today it’s time to value the Australian Equities market (proxy ASX200 CAPE) where I currently hold 18.9% of my net worth.

As of the close on the 12 May 2011 the ASX200 stood at 4696.  Based on this price the first chart today shows the cyclically adjusted PE (ASX200 PE10 or CAPE) at 16.6 which is well down from the 17.8 when I last ran the numbers back in December.  The long run average since 1993 is 22.5.  Is this implying the ASX200 is undervalued by 26%?  I’m not sure but what it is saying is that my target tactical allocation is now 20.2%.  If you’re interested in how my strategy was built then a good place to start is here.  There is also a lot of information in the sidebar links.  The CAPE value is also not far from April’s PE value of 17.

As always my 2nd chart shows the ASX 200 PE10 versus the 1 year real return that can be expected.  The correlation is currently -0.36 which is one of the main reasons I use the ASX200 PE10 as a metric for the tactical portion of my asset allocation.  It suggests over the long term I might be able to squeeze a small amount of extra performance from my portfolio.  Using the trend line of this second chart with an ASX200 CAPE of 16.6 shows an expected 1 year real (after inflation) return of 14.3%.

Additional CAPE/PE10 data for this month includes:
- the correlation between the ASX200 Price and the ASX200 CAPE is currently 0.75. 
- the ASX200 PE10 average for my dataset as mentioned above is 22.5
- the ASX 200 PE10 20 Percentile for my dataset is 17.2
- the ASX 200 PE10 80 Percentile for my dataset is 27.6

Today’s 3rd chart today shows Real (after inflation) Earnings and Real Dividends for the ASX200.  Dividends are at 196.3 which is still below the long term trendline.  Earnings on the other hand are now just about on trend at 283.7.  Why are companies hanging on to all their earnings?  Have they learnt lessons from the last downturn and want to be prepared for the next one or can they see the next downturn approaching?

As always do your own research.

Assumptions include:
-          -All historic figures are taken from official data from the Reserve Bank of Australia.
-          -Latest P/E ratio is the April 2011 ratio taken from the RBA official data
-          -May 2010 price is the 12 May 2011 market close of 4696.
-          -May 2011 Earnings and Dividends are assumed to be the same as the April numbers.
-          -Inflation data from April and May 2011 are extrapolated.


  1. Good to have you back on line - I missed your posts.

    I'd like to understand how you calculate the ASX PE10...

    I gather price is the index value which you take from the RBA site but could come from a number of sites. Right?

    And I also understand you get inflation from the Australian Bureau of Statistics. Right?

    But where do you get earnings from? Do you back calculate from the PE(1) data ( at the RBA?

    Are you aware of anyone else publishing PE10 data for the ASX?

    Best wishes with settling back into UK.


  2. Hi TF

    Thanks for the wishes. Life is settling down in the UK. All I need now is a job and normality will have resumed and I can enjoy the summer.

    I think you have my method for the ASX 200 PE10 spot on. Just for clarity I will detail step by step to detail the inflation calcs etc:

    1. Price is taken from the RBA site that you quote above with the exception of the last data point that I usually quote which is just usually a close of trade price on the day I publish my PE10. I always quote what and when that Price is in my post.
    2. Price is then corrected for the CPI from the ABS. The last few CPI data points are just extrapolations from the last 2 known CPI points.
    3. Earnings are exactly how you say. Earnings=Price/PE taken from the RBA. The last earnings point is just the earnings from the previous month.
    4. Earnings are then corrected for the CPI in the same way as Price.
    5. PE10 is then the Real Price / Average 10 Year Real Earnings.

    Nobody that I know is publishing data for either the ASX200 or the FTSE100. I spent a lot of time looking. (If you know of anybody it would be great as accuracy could then be compared.) That's why I built the datasets which took me a looooong time as a lot of it had to be done manually. The only concerns I have are that both datasets are quite small and nothing like the US dataset going back to 1881.


  3. Absolutely superb analysis. My only criticism is that it would be of greater value to see the PE10 analysis with data going back a lot further such as to 1890 give a better perspective. I presume the forward one year return series uses longer data but the data range was not shown in the title. The PE10 is much more predictive of 10 year forward returns than the 1 year return as I imagine you understand well. A factor to consider is the extent to which earnings have increased on a temporary basis owing to significant increase in aggregate debt over the last 30 years which is a one-time event - an aggressive de-leveraging process has only just started which will take quite a number of years go end, harming earnings and sentiment for some time.

  4. Hi Julian

    Thanks for the compliment.

    I fully agree that it would be great to have data going further back in time. I make this type of comment regularly in my posts. At the moment I have S&P500 PE10 data going back to 1881, ASX200 back to 1993 and FTSE100 back to 1993 (albeit with an averaging fudge for the 1st 10 years). You wouldn't believe how hard or how long it took to just build the data I have for the FTSE100 and ASX200. Unfortunately it's the best I can do with the resources I currently have available to me.

    The forward one year return is based on monthly data and uses exactly the same time period as the other charts.

    I haven't run a forward 10 year returns for the ASX200 or FTSE100 as I feel my dataset is to smll. I have however in the past run an analysis in the past for the S&P500 PE10 vs the Real (after inflation) Return using annual data. The results match what you say, which also seems logical to me. The correlations were:
    PE10 to 1 Year Real Return -0.21
    PE10 to 5 Year Real Return -0.41
    PE10 to 10 Year Real Return -0.52

    I've also postulated in some posts previously about whether the earnings recently are indeed temporarily high due to debt loading so I hear you on that one. It is certainly a consideration.


  5. Hi RIT,

    Indeed, earnings in both the US and Australia are presently quite
    above the sustainable level calculated using any of the varieties of
    on-trend methods. This is putting the problem of debt leveraging

    Also the US market was indeed very overvalued on average for most of
    the last 25 years compared to long run average multiples of net
    earnings, and few people invest actively for much longer than that, so
    most investors think this is the normal range.

    This chart may also interest you.

    The following is an extract from Buffett's fortune article disussing
    the market valuation and a link to an updated Market Cap as a % of
    Nominal GDP:

    "On a macro basis, quantification doesn't have to be complicated at
    all. Below is a chart (similar updated chart by thechartstore),
    starting almost 80 years ago and really quite fundamental in what it
    says. The chart shows the market value of all publicly traded
    securities as a percentage of the country's business--that is, as a
    percentage of GNP. The ratio has certain limitations in telling you
    what you need to know. Still, it is probably the best single measure
    of where valuations stand at any given moment. And as you can see,
    nearly [11] years ago the ratio rose to an unprecedented level. That
    should have been a very strong warning signal. "

    I wouldn't pay a great deal of attention to that however, except when
    we are at extremes (such as 50% or 170% of GDP). It does not factor
    profits rising outside of the US among other problems with the model.
    The *only* way to look at the market valuation for the perspective of
    projecting long-term returns (ironically the only term that is
    amazingly well forecast-able) is to look at the on-trend present
    earnings compared to the average on-trend earnings over the last
    century. On that basis the US market is about 35% overvalued now and
    should get a real return somewhere between -1% and 3% over the next 20
    years, probably in the middle of that range and very little chance of
    being out of the range. As for what happens over the next 2 years
    through, almost anything - there is absolutely no way to forecast.
    Very short-term there is a lot of momentum and recent highs recently
    being hit which is predictive of further short-term good performance -
    but that eventually comes to an end.

  6. To get a higher long-return one needs to be allocating capital around
    to areas of lower valuations on a continual basis as you are doing.
    Another way to do this to put the money in the hands of a corporation
    that is focused on capital allocation and not merely generating
    profits without great skill of where to invest new capital. Examples
    are LUC, BRK, WRB. BRK is presently particularly cheap (117,000) so
    you have the advantage of long-term growth in intrinsic value
    exceeding the market - amplified by the eventual return to the company
    trading at price/book multiples closer to the historic range. You need
    to model in your mind the two things separately - how a company is
    able to grow intrinsic value, and secondly the market price's
    relationship to intrinsic value - so you want something that grows
    intrinsic value faster than the market with more or less certainty,
    then amplify your returns by finding the companies also trading below
    intrinsic value. Berkshire is almost certain to outperform the market
    over the next 20 years by 1% to 5% (if valuation was presently normal)
    and 3% to 8% based on the current low valuation.

    I'm still doing more research to work out to what extent (if any) the
    ASX 200 is overvalued. Considering that house prices have yet to start
    falling but have a terribly long way to fall *and* the aggregate debt
    leveraging having accelerated downwards aggressively last year (that
    is 'acceleration' downwards, not velocity - so public+private debt has
    just started to move downwards very fast but it still very high so
    no-one is complain that much yet), the public are not really
    appreciating how difficult this is to stop. This is also consistent
    with my preliminary research showing the ASX 200's aggregate earnings
    to be above trend (made possible by aggregate debt at or very close to
    rare maxima).


  7. congrats on your site, you are ahead of most pros and your thinking is sound IMO.
    this data source might be of interest to a curious mind like yours in case you haven't seen it.
    i've cross checked some of the data and it looks legit, the site's author is well known here in Aus.
    various PE series for Aus do differ (some are incl lossmakers, others excl.)