When to buy & sell housing

Fellow econblogger, Cameron Murray, has written another fascinating post on his Blog about the best time to buy and sell real estate. Cameron’s post is re-produced below for your reading pleasure. As always, comments are welcome.

I came across the Commonwealth Bank – RP Data Home Buyers Index recently. It is designed to estimate the balance of supply and demand in a suburb to indicate whether it is currently a ‘buyers market’ or a ‘sellers market’. Their website explains:

The Commonwealth Bank – RP Data Home Buyers Index estimates effective supply levels based on the number of properties being advertised for sale within the region.

…On the demand side of the equation, Australia’s largest home loan lender, the Commonwealth Bank, provides a summary of the number of home loans that have been funded across Australia. Once we factor the Commonwealth Banks share of market into the equation, the number of home loans funded provides one of the timeliest estimates of housing demand in the market place.

This indicator may signal which direction prices are moving at any point in time, and is therefore a useful tool for market analysts. However, I was wondering if there is a rule of thumb that residential property investors could use to time their entry and exit from the market to maximise returns?

To answer that question I propose Murray’s Retrospective Indicator for Buying and Selling.

Fundamentally the returns from property are rents. Capital gains are a secondary return primarily resulting from changes in macro monetary conditions, with additional gains from changes to density and land use. So without any data at hand I would say a good rule of thumb is to buy yields and sell capital gains.

But creating a workable measure from that requires recognition of the economy wide conditions at the time, particularly regarding interest rates and the relative strength of yields in comparison.

So my simple measure is the mortgage rate divided by the market gross yield. When this measure hits 1.5 it is time to buy. When it reaches 2.3, it is time to sell.

Let’s look at how you would have fared investing with this indicator since the 1980s.

First, a graph of the mortgage rates and the market gross yields over time (from Chris Joye):

When I first saw this graph it was clear to me that the relationship between these two rates was a good indicator of the market. Put simply, when the lines are close together, yields are getting close to matching the mortgage rate. The further apart they are, the lower the yield compared to the price of money. This means that prices may have climbed beyond what is justifiable by yields, with capital gains only able to return again with a lower interest rate setting.

That translates to a ‘buy on yield’ signal when the lines are close, and a ‘sell on capital gains’ signal when the lines are further apart. So I plotted the ratio of these variables to show how the rule plays out.

As I said, the buy signal is when the mortgage rate is a lower proportion of yield (1.5x the yield) and the sell signal is when the mortgage rate is relatively high (2.3x the yield).

An investor following this rule would have had two holding periods – early 1983 to mid 1988, and late 1996 to mid 2003.

Over the whole period, the average capital gain was 7.3% according the ABS capital city established house price index (the index starts in 1986). We can compare these two periods’ returns to this average.

We will say that the first holding period is from mid 1986, since that is when the ABS price data is first available and where the buy line is hit for the second time, to a selling time in Sept 1988. The capital gain over this period was 16.2%pa, and the yield at purchase was 9.5%, giving a 25.7%pa gross yield over the period. Mortgage rates over this period averaged 14.7%, so leveraging would have greatly increased the return on equity.

The second period goes from Sept 1996 to Sept 2003. The capital gain over that holding period is 10.4%pa with a gross yield at purchase of 5.3%. This gives a 15.7%pa gross return over this period (again, much more with a little leverage with mortgage interest rates averaging 7.1%). Had you held off buying until the indicator left the buy signal in 1998, your capital gain over the 1998-2003 period would have been 13.1%.

The rule might even work with the latest swing in the market. Late 2008 was by all accounts a good time to buy, and assuming the indicator hits the sell line during 2010, that would have been a great time to sell. In this period you would have made capital gains around 9%pa according to ABS figures, or 7% from Rismark figures, with yields at purchase around 4%, giving a 13% (or 11%) annual gross return (see dwelling price chart below).

Had you followed the opposite pattern, and bought on the sell indicator, and sold on the buy indicator, your returns would have been far from impressive.

In the period between 1988 and 1996 the capital gain was 4.3%pa with a purchase yield of 6%, giving a 10.3% gross return. The mortgage interest rate over the period was 11.8% on average, so gearing would have only reduced these returns.

In the period 2003 to 2008, the capital gain was 5.3%pa, and purchase yields at 3.5%, giving 8.8% gross returns. The mortgage interest rate over this period averaged 7.6%, so there were additional gains from leveraging.

The table below summarises the periods signalled by this indicator (with the green columns the buy and hold periods, with 2008-10 using the more conservative Rismark index for capital gains).

So it seem that this magic indicator would have maximised returns for property investors if we ignore the costs of buying and selling (which may offset much of the gains from short term holding periods, and lead to longer term holding periods generating better real life returns).

Ideally one should examine the effectiveness of the indicator at a city level, since the timing of property growth cycles in each city is slightly different, and the buy/sell signals will be slightly different. I would be happy for any data provider to use this indicator to do just that.

I have called it a retrospective indicator because it has only worked in the past three decades where government have been able to ratchet down interest rates in the face of crisis. Given that the scope for further serious cuts in mortgage interest rates is diminishing the closer we get to zero, and with more risk factored into lending rates, I wonder how much longer these trends can keep going.

My guess is that we are currently still sitting close to the sell signal. Yields appear to be firming as prices slide, perhaps edging the indicator down below the 2 mark. Assuming mortgage rates stay constant, and rents are flat, to get back to the buy indicator prices would need to fall by approximately 30%.

But if mortgage rates fall 2% following action by the RBA in the coming years (which would need more than a 2% drop in the cash rate), prices need only fall 6% in order for the indicator to reach the buy target (from Dec 2010 levels). So it is likely that reality will be somewhere in between.

So don’t buy yet.

*I have used the ABS capital city established house price index to estimate capital gains, except for the 2008-2010 period, where the capital gains are estimated from charts of the Rismark index (which is a better measure in my opinion). The yield data is obviously generated from another data series, so the numbers generated are indicative only at a national level. I expect the pattern to hold with consistent data use, and on a city level, although the buy/sell signals will be at slightly different ratios.

Unconventional Economist
Latest posts by Unconventional Economist (see all)


  1. This is great… I like this indicator much better than the median price v (disposable) income matrix.

    I wonder how Houston and Munich fare?

    • Exactly…..!

      This formula really only applies to volatile markets, which in turn, happen to be “growth controlled” ones.

      In Houston and Munich, you only “invest” in a house for honest, safe, rental returns. You don’t expect capital gains to make or break your whole investment; or “picking the timing” during a volatile “cycle”.

      I recomment the books on picking timing, by Kieran Trass; I think the latest one is called “The Housing Bubble”.

      He fails to pick the central role of anti-growth regulations in increasing the volatility of these cycles, but he goes into the sort of thing Cameron Murray is talking about, and a lot more contributing factors too.

  2. Something is wrong with the commbank / Rpdata index . I try out an FHB area of Melbourne and the index shows it as a sellers market (sellers having an upper hand). However the Rp data house price index in that period for that suburb is showing house price falls. Also I have anecdotal evidence of people offering below the asking price.

    • Woh! another suburb which has fallen close to 10% according RP Data comes up as a Sellers Market in the index

  3. This seems silly…basing long term buy decisions on short term fluctuations between rental yields and the price of money as set by the RBA…

    Surely you would find yourself buying and selling into the market every time the RBA overshoots one side too far…and this is not to mention all the transaction costs.

    I think buying with a belief that the RBA will slash rates and give you a good yield is a silly way to invest. I dont want my investments contigent on the policy of the RBA

    • Maybe it’s better as a contra tool? An indication of ‘when NOT to buy or sell’ if that’s part of a descision making process.

      • Agree Janet…that seems a far more useful way to use this indicator. Because making long-term Buy decisions based on the premise Cameron describes is setting someone up for long-term pain.

        This isnt share trading…housing is not somethng you can get in and out of with ease so this kind of analysis is flawed/pointless.

        Sorry – I really like what Murray writes usually and maybe I have missed the point here?

        • He does note that it ignores the costs of buying and selling….

          I guess you’d want to be satisfied that interest rates were going to be relatively stable for a period in the future (rather than just a short term drop like in 2009).

      • ceteris paribus

        Brilliant point Janet. It is one very worthy contra-indicator for a buying and/or selling decision.

        Another point. I am a bit sceptical about “trading” property- but I am not convinced the author is even suggesting that in his blog. He is just giving us one important evidence-based insight into the bigger timing puzzle.

    • Good point Stavros. I was thinking the same thing. You would only use the buy signal if the interest rates were higher/more relaistic.

  4. Great piece of correlation indentification between rent, rates and prices. But if credit availability decreases that correlation will disappear.

    Lastly, isn’t there enough analysis on this blog to show that housing should not be treated as an investment? Showing historical returns on an asset which depends on future generations taking on the sellers debt would seem to head into a wall at some point.

  5. For an average owner-occupier:
    Buy when you need new home to live but only if you are expecting to live in that home for a longer period of time (>10 years) and in addition home price is not inflated in respect to alternatives (renting).

    Sell only when you have to move (to another place that makes commuting impossible or time very long). Do not buy new home in new area unless you are quite positive you will stay in that area for longer period.

    For investors:
    Buy if you have cash to buy (buying on credit is usually not profitable, and even if it is at some point of time it will not last for long); in addition: if you can get decent rental return (co cover ongoing costs and provide decent profit) and if realistic price increase over longer period (10+ years) can cover inflation and fixed buying/selling costs.

    For “investors” (speculators):
    Buy whenever you want. Gambling is addiction that provides some satisfaction so go ahead.

    For others:
    Large fixed costs of buying and selling make home trading really expensive hobby.

    • all this other crap about seller’s or buyer’s market; “predictions” about price movements, profits, NG, currency movements, etc. is bullsh**

      if you want to earn money on housing start construction business and create something that society may benefit from

  6. +100 excellent article.

    Out of interest is it possible to show the mortgage rate/gross yield chart for capital cities – in particular Brisbane?

    Thanks and keep up the great posts

  7. Your system would have resulted in your selling in 1988 and missing out on the enormous 40% capital gain over the next 2 years (ABS index went from 70 in March 1988 to 100 in March 1990). That 40% rise from 1988 to 1990 is confirmed by Stapledon’s data.

    Anyway, your work does at least dispel that myth that it used to be cheaper or the same to buy as renting. People still say they’ll buy when its cheaper to buy (just comparing rent to repayments in the first year).

  8. Thanks for the thoughtful comments all.

    The numbers come from a sell indicator point in Sept 1988 on the ABS series, as the indicator crosses around July. The reality would be that selling would take some months from when you know the indicator has reached the trigger. But you do miss out on some massive growth immediately after.

    It does certainly dispel the myth that it used to be cheaper to buy than rent on average.

    But these are average figures and I know that in some areas the ratios would be much lower. The buy trigger might be at <1x in these areas, meaning that interest payments on 100% of the price are less than the rent at that point in time. Net yields (after other costs of ownership) exceeding rents would have been historically and geographically rare. That's why I believe it would be a very interesting exercise to do this analysis at a city level.

    Also, from my understanding, the yields are a simple average rent/average price, but we know that the rental market is typically a lower value subset of the whole market, so the actual yields, it just for the rental market homes, might be a fraction higher.

    And of course they are gross yields.

    Most interesting was that the period 2003-2008 was the most expensive (relative to rents and prevailing market interest rates) since the 1988-1991 period. Indeed, for about half the study period homes were around 20-30% cheaper (relatively).

    The indicator is simple rule of thumb about when has been a relatively good or bad time to buy and sell for the investor. OF course the transaction costs are high and I acknowledge that it might be better to hold rather than sell when the indicator suggests.

    But surely using this measure to time entry to the market, even if you are investing over a 10+ year horizon is superior to, well, just guessing? If you had bought in 1990 it would have taken years just to recover your costs, let alone match returns from keeping you money in the bank. And if you were leveraged, the situation was even worse. Even selling after 10 years, in many places you may not have broken even. And would be unlucky enough to find that 3 years later the property you sold had doubled in value (or is that price?).

    And as others have suggested, it also works well as an indicator of when not to buy or sell. As I said early in the post, the indicator tells you when market prices are relying more on capital gains for their perceived value, or relying more on the rents. You want to buy as an investor for the rental return, with any capital gains a bonus.

  9. “And as others have suggested, it also works well as an indicator of when not to buy or sell.”
    It certainly does not work as a good indicator of when not to buy. It indicates 1994-96, 2005-2008 as times not to buy. They were very good times to buy.
    The ABS 8 cap city index has risen 45% since 2005, 39% since 2006, 27% since 2007 and 17% since the end of 2008. Even buying in 2009 was good.
    But these are averages for the cap cities. Melbourne prices have risen 70% since 2005 which the indicator says was a time not to buy.

    • ? I’m not sure what you are trying to say now Suzi.

      Yes in the long run prices have risen, but your total returns would have been better if you bought when this indicator suggested – I thought that was what the figures were saying.

      2008 was a terrible time to buy – the capital city price index fell 5.5% in a year from march qtr 2008. Therefore had you waited till 2009 you would be much better off both in 2008 and now.

      And values only gained 1.2%pa from 1994-1997, so there was nothing to be gained from getting in at the beginning of that period, given the 10% interest rates and 7.2% gross (rent+capital gains) yields.

      In 2003-4 yields were around 3% with interest rates around 7%, so anyone with leverage would have needed quite a deal of capital growth just to break even, let alone make a decent return. You also need to remember that when you sell out of property when the indicator suggests, you can then invest your money in other sectors. The share market from 2003-2008 was booming.

      When the indicator is above the sell line that does not mean prices will not continue to rise, just that they are more likely to fall or be flat, since there is little room for yields to fall further relative to interest rates.

      So for you it was always a good time to buy? Surely though, some times are better than others?

      • Obviously there has to be a point far too near the TOP of a volatile house price cycle, to “buy”. I like Cameron’s theory of “when” that point is.

        The world is full of idiot property investors who have taken massive haircuts and gone bankrupt big time, because they bought too near the top and got caught by a following big dip. Even if they hung on without going bankrupt, it has been YEARS before some of them got back into positive equity territory at all.

        Why anyone expects Australia to not do the same to its property investment sector, is beyond me. I might have an explanation one day, for how they avoided a California/Ireland/Spain style meltdown – if they do. Right now, Australia is baffling me with its “Indian Rope Trick” property price graph.

  10. MAYBE, it is something to do with the Australian Federal Reserve being able to do things “BEFORE the crash”, that were not done in Ireland, Spain, and California because they are “member states” of a much larger monetary/political union.

    Even so, I remain totally unconvinced that this is wise. I think Ludwig Von Mises was right:

    “There is no means of avoiding the final collapse of a boom brought about by credit expansion”

  11. Plus, I cannot see that an economy can stand an indefinite trend towards:

    Higher urban land costs than its international competitors.
    Higher labour force “housing” cost inputs than its international competitors.
    Lower household discretionary incomes than its international competitors.
    Land-intensive industries squeezed. Labour productivity reduced.
    “Unintended consequences” for urban “form”, of inflated land prices (location decision signals swamped by land prices, leading to sub-optimal location decisions regarding transport, agglomeration efficiencies, proximity to amenities, etc).

  12. Actually – this is how buy a house…haha


    Investment in research has first-timers sitting pretty

    Property developer Aiden Werry, 25, and his fiancee, Michelle Evans, 25, had been looking for 18 months before they secured their two-bedroom apartment in Marrickville for $605,000 through McGrath. It last sold for $520,000 in 2007.

    ”I had been living at home so there was no real rush,” Werry says. ”It gave me time to get familiar with the market, to understand prices.” He says he much prefers buying rather than renting. ”I’d rather pay a mortgage than throw money at a landlord for no real return … we couldn’t really afford a house near the city but thought it would be nice to have an asset or equity.”

    When the couple became serious about buying somewhere in January, they found that they were being beaten at auctions.

    ”So when we saw this one, we bought before auction,” Werry says.

    To first-home buyers currently in the market, he advises against rushing and says to go to plenty of inspections. ”That way when you do decide to buy, you can make a move, an educated move.”