Measures of Affordability in Context. And Is This Chris Joye’s “Fischer Moment”? (Updated)

In case you didn’t know, my favourite property spruiker, Chris Joye (CJ), has recently posted this article in Business Spectator:

A Property Bubble Long Shot

More of the same: ‘we’re right, they’re wrong, Australia is different’:

In one corner we have journalists at The Economist newspaper, who in their latest survey make the extraordinary claim that Australian house prices are overvalued by 56 per cent using their preferred benchmark. In the other corner we have a crowd of the most respected economic minds in Australia, including the Reserve Bank of Australia, the Commonwealth Treasury, almost all market economists, and leading house price index providers, such as RP Data and Rismark.

In my opinion, this is not a goo start – it rings of an argument from authority, which I have little logical respect for. Anyhow, moving on…

I should add, there is an implicit (and under-recognised, IMHO) assumption here, that certain measures of prosperity and affordability–point-values, such as ‘median this’, and ‘average that’–are actually reliable indicators, with little acknowledgement of the reality that such measures are, technically, ‘distributions’ or ‘curves’ (to use engineering-speak), not just points in space that necessarily adequately reflect the reality of the entire statistical population.

No, in fact, such point-value measures do not reflect reality for many; most distributions are terribly skewed to one side, making them too simplistic and unrepresentative to assess broad and diverse issues such as housing affordability.

Given this, the zeal, confidence and emphasis with which CJ et al tout these measures is concerning, especially considering the influence these people have with fellow economists, government bodies, financial institutions and consumers, and the way these measures are uncritically received.

CJ et al assert, referencing their measures, that Australian housing is not nearly as over-valued as some attest, thus we are not trapped in an ever-expanding property bubble at risk of imploding. Yet they concede in other articles that the measures they use are, indeed, historically high, or at least a significant problem warranting significant action!

And if this is the case, why are CJ et al so confident?  Is there sufficient contingency in many households’ budgets to handle increases in the net cost of managing a mortgage? Is there any fat left to cut off? Any buffer? And if one asserts that there is, where in the distribution is all this “fat” located, and how well does it represent the population?

Yet the truth is that the point-measures are given far too much weight, and now, as peripheries at the lower end – that represents most of the population – are struggling, these same measures are still being interpreted as “everything is fine (but still expensive…but no worries!). And the peripheries are what can (and do) bring entire systems down – like a burst puss-filled abscess, or a tumour turned malignant. The US sub-prime mortgage debacle was just one such example, in that, even though their actual fraction of total mortgages was “relatively low”, it brought down the entire system – and some say, the entire economy.

Let me say this: look, all of you, bull or bear, for crying out loud we have to stop getting caught up in the details of which measure we are considering.

Please allow me to throw some practical engineering-type thinking at this (as is my personal angle): different methods of measurement and expression will yield different values and different calibrations; but, because they are, essentially, measuring much the same thing – wealth and housing affordability – they are simply giving similar results, just with different base calibrations. There are differences – sure, but none of them are essentially, philosophically, different enough to really say too much more than the others – whether median this, average that, ratio of what and whatever. They are all too similar philosophically for one to say much more than another.

Hence, whether when we use a certain measure, it is only relevant to itself, and should not even be compared to the results of other methods.

To break it down a little more: consider the notion I will call “Absolute Relevence” – or shall I call it “Intra-Relevence” – can only come from comparing the results of the SAME method. This would be, for example, comparing average incomes to average prices, generating results, then comparing the results OF THAT VERY SAME METHOD between cities and countries.

Now consider: “Relative Relevance” – “Inter-Relevance” comes from comparing the relative changes (eg. %) of Absolute Relevance values, for example, in an indexed form (a common economic method). However, comparison of the absolute values is totally irrelevant and simply philosophically, logcially and statistically invalid.

Those absolute values, though they are measuring essentially the same phenomena, are derived from different methods, coming from different philosophical axioms – direct comparison is totally invalid, and should be discarded. But relative comparison, as mentioned above, is a different kettle of fish, and lends itself to far greater levels of logical validity and soundness…and, hence, reliable, real world usefulness.

Hence, taking the comfort they seem to from confidence in their own method – which is fine – but then outright dismissing that of The Economist deserves, IMHO, the following treatment:

1) It has some validity – They have somewhat addressed the axiomatic philosophy of The Economist publication; hence, if they disagree with its axiomatic foundations, then dismissal of its outcomes is somewhat valid (though it does not necessarily follow that the outcomes of the other method(s) are entirely wrong and unreliable – that is another pursuit entirely)

2) The actual stated application(s) that CJ et al express in their many mainstream articles and addresses, of their own results, is invalid – CJ et al derive numbers from their own method, then compare those numbers to the numbers of The Economist (Steve Keen, etc, etc). Hence, they imply, because the CJ et al numbers are lower than the numbers of those using different methods, then housing is not as unaffordable as we all think. And this despite the admission of CJ et al, that housing is expensive, or at least as expensive at other expensive places in the world, as that we should somehow be OK with that. When the reality, I argue, is that they are different methods yielding different numbers that “mean” essentialy the same thing.

But, nonetheless, this an absolute comparision between values of derived from different methods is fundamentally flawed.

No, instead, as I discussed before – the numbers have no absolute relevance to eachother: ie. a “4.5” from one method has no absolute comparative relevant to a “9.6” from another method. For example, it is invalid to say that that a method, say, measures housing affordability for Sydney, Australia, and generates a 9.6 value, should be considered to be in the order of two times over-estimated because adherents of another method got a figure of, instead 4.5. That is a totally invalid and irrelevent conclusion.

Instead, the 4.5 is a description of a certain real world affordability from one point of view; and the 9.6 is a description of much the same certain real world affordability from another point of view.

Hence in that sense, the 4.5 ~ (approximately equal to) 9.6, but just analogously linked, and differentiated by different (but still quite philosophically similar) axioms. Given their philosophical similarity, an engineer (or whoever!) would quite validly – and even just empirically – determine a calibration function/algorithm to link the two methods. That is, the value of one affordability from one method could be estimated by knowing the value of another method, via processing through a function/algorithm.

So…why is this possible? Because they’re philosophically similar and have very similar philosophical roots, modes and objectives, and are concerned with much the same real-world phenomena: affordability.

And why would someone even consider something like that (not that i’m suggesting it SHOULD be done, just that it could be done)? Because they are different methods giving different numbers, which, internally and relevant to themselves essentially say the same things in a different way.

And in the specific case of Australian housing affordability, when they observe essentially the same general phenomena, 4.5 (or whatever) is “expensive”; 9.6 is also “expensive” – so there is, therefore, a justifiable correlation between the two numbers, but they are only relatively comparable, via something like an index, and not absolutely comparable.

QED?

Therefore, please: stop it, bull or bear. Stop the comparisons of numbers from different methods. Just stop it.

“Expensive” is, truly, a qualitative notion, and should be viewed in that context FIRST.

That is: ask people how they’re going financially; ask them how far their wage goes now compared to another time. Get a gist: believe it or not, it’s actually the proper context to understand all the magic numbers coming out of all the various methods of wealth and housing affordability. Then you will know what your numbers “mean” in the context of the real world they are attempting to describe.

And what do we find? People are doing it tough. Very tough. AND, what’s more, LOTS of people are doing it very tough.

Got the context now?

So, what is the context and implication of the significance of the 4.5 from one Method 1, the 9.6 from Method 2, the XYZ value from Method 3? They all relative indications of “expensive”, and “tough”, and “maxxed out” and “unaffordable”…insert your favourite adjective.

Perhaps I am being a little harsh…but I am somewhat frustrated at the faulty logic being thrown around, either by bull or bear. In truth, I think CJ is a bright spark – at least that’s how it seems. And, boy, does he have his hads on data I wish I could get my hands on!

Though not explicitly featured in the above-referenced article, CJ et al have been (and still are?) comparing their 4.5(ish) (or whatever) numbers to someone else’s 6.0, 7.5 or 9.6 or whatever; and at least implying, if not stating outright, that since 4.5 is less than XYZ, then things are not as bad as you all think.

Invalid. Irrelevant comparison. Throw it out. Start again.

But instead we get an exploration into a failed replication of The Economist’s methods, an assertion that 38.7% is basically the same as 56%…

If we adopt The Economist’s method, we conclude that the current ratio of median Australian house prices to median rents is about 38.7 per cent above its 28 year average. We do not know how The Economist gets its 56 per cent estimate, but ours is in the same general ball park.

Really? Hmmm…not too happy with that assertion myself.

…and THAT, after we are told:

In order to decipher who is right or wrong here, one has to dive into the detail.

However, in light of my earlier, more philosophical breakdown of the matter generally, it is NOT about the detail, it is about axiomatic starting points; because, when comparing paradigms, one tend to get LOST in details, not enlightened by exploring them – the problem is axiomatic, not semantic; logical, not mechanical. Don’t step in, step out. We’ll see that they’re all similar measures of “expensive” and “unsutainable”, as that is what so many people are telling us they are experiencing now, and much more than that previous. There’s the calibration context!

And now I come to my two-point finale:

Firstly, more of this sort of argument – international comparisons of a various measure of affordability, via statements like this:

…dwelling price-to-income ratios remain unexceptional by international standards.

What other nations? USA? Britain? Various European countries? Other Asian countries? I’m not asking them to be specific, but, instead, I would be quick to point out that many of those nations are either in, or close to being in recession, depression (think Ireland…); or having very significant and unsustainable inflation problems due to ridiculous credit booms?

If I analogously compare high housing prices and unsustainable economic models to a sickness, then should a sick man feel better because he is not as sick as someone that is very sick or dying? Perhaps – but they are still both sick, and that’s not “good”, or even “OK”, right?

And who is to say that the sick man does not have what the dying man has, just because they are offset in time? One should be cautious and reflective, not assume “I’m different”. Hardly comforting, really. What if we’re all just going down together? How would you know? Maybe time to stop comparing ourselves to the sick men as if it were a good thing, eh?

Secondly, the following two quotes made me think “plateau speak”:

In the other corner we have a crowd of the most respected economic minds in Australia, including the Reserve Bank of Australia, the Commonwealth Treasury, almost all market economists, and leading house price index providers, such as RP Data and Rismark.

This latter cohort essentially contend that The Economist does not know what it is talking about. They argue that Australian house prices are not materially overvalued, and there is no reason to believe that they must suffer precipitous price falls in order to obtain some more desirable valuation benchmark.

And:

Looking ahead, it is highly likely that Australian house prices will track household earnings in what PIMCO’s Bill Gross has aptly described as the ‘New Normal’.

…and when i think of “plateau”- and “new normal”-type-speak, I cannot help myself to think of the famous quote of Irving Fischer, not long before the famous 1929 stock market crash (which led to the Great Depression):

Stock prices have reached what looks like a permanently high plateau.

Hmmm….

Now understand that I’m not saying the CJ et al are saying prices won’t go down; no, he’s asserting that they are largely tied to wages, which to a large extent, I can agree with; however, what I DISAGREE with is that they will be tied to wages AT THE SAME LEVELS/RATIOS THAT THEY ARE TODAY, as if that were both desirable and sustainable for the populace, and that they will not decide and act contrarily.

But is this CJ’s “Fischer Moment”? Is the Australian Housing market turning down in both nominal and real (inflation-adjusted) terms at this very moment?

I personally think it “turned” in October 2010 – but I, admittedly have a particular way of looking at and interpreting dynamic data, so others may not agree with quite the same gusto, even if they are bearish.

Truly, there is a lot of detail in that piece by CJ – but the relevance of the detail must be understood in the context of one method-adherent speaking to another method-adherent, separated by axioms, but essentially measuring and describing the same things: wealth and housing affordability.

In fact, one would be better considering if a suite of measurement methods would be a better approach, and seeing how they all change throughout time, or if one if concerned with detecting problems. But none can be considered without proper qualitative consideration of the bigger, realer world, including the economy of the man on the street, and his little average house and family.

But being constantly reminded that we should be comforted, because a certain number is less than another number from a different method, is invalid at best.

Regards,

Stewart

Advertisements

4 Responses to Measures of Affordability in Context. And Is This Chris Joye’s “Fischer Moment”? (Updated)

  1. Ron says:

    One of the most coherent articles I’ve ever read on the topic of housing affordability measures. Individual people have their own anecdotal evidence that they hold truer than any number an economist or uber-bear/bull spits out at them. My anecdotal evidence based on my experience and that of my friends, family, extended family, co-workers and acquaintances tells me that house prices have sky-rocketed in a short period of time. Only time will tell if the winners of this period will be the winners of the next ten to twenty years or if those who “missed the boat” have indeed escaped a precarious financial situation and will reap the rewards later down the line.

    • Stewart says:

      Ron, thanks for your kind words. And your experiences echo those of my own. Hopefully this all ends “nicely”, but I wish I could realistically be optimistic, but I can’t…

      See you round,
      Stewart

  2. Timbo says:

    Really great article.

    Your arguments can apply to many markets, not just housing. There is an over-reliance, especially by economists, to draw way too many conclusions from their models. We are dealing with systems here, complex systems to be precise, and the simplification of these to a few numbers and a linear equation is one of the reasons countries around the world are in this mess to begin with. It is no coincidence that engineers, physicists and computer scientists seem to make up a bulk of the critics of these models. Not only are they used to systems thinking, but they are mostly perplexed by the over-simplification seen in many economic models. A lack of formal training in economics seems to be an advantage in times like these.

    Cheers,
    Timbo

    • Stewart says:

      Timbo, can’t add anything to your comments.

      Thanks for visiting!

      Stewart

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: