Banks, statistics and the business of surprises

Banks should have a good handle on how the economy is travelling. Banks should have a good handle on how the economy is travelling.
Nanjing Night Net

Banks should have a good handle on how the economy is travelling.

Banks should have a good handle on how the economy is travelling.

There was bemusement to be had on two fronts yesterday concerning Australia’s banks: the “surprise” of their share prices bouncing; and the “surprise” of lower than expected retail sales numbers.

In the latter case, if the big banks’ chief information officers are worth their large salaries, there should have been no surprise at all in the August sales figures.

Or the Australian Bureau of Statistics has had another bad month, seasonally adjusted.

In the former, the game of selling off bank shares after they reach record highs as various analysts talk them down, only to see a subsequent bounce, is becoming as seasonally reliable as “killer toys” stories a couple of months out from Christmas and a pick-up in interest in horse racing in early November.

The commentariat and bank analysts’ cycle goes something like this: talk up bank stocks until they reach record highs, then talk down bank stocks as being over-priced, then, after they have a little correction, start talking them up again until they reach another record high. Repeat.

The supposed problems with the major banks since they released their results – alleged poor growth prospects, alleged poor bad debt outlooks given the low level of bad debts, alleged debilitating new capital requirements as regulations tighten and the alleged end of yield-chasing – are all rather relative.

Knock a few percentage points off the record share prices and suddenly those problems aren’t so bad.

So the game resets and away we go again.

Yesterday’s action reportedly was led by foreign buying interest that was helped by the Australian dollar’s fall.

That could be an interesting factor – some big foreign players punting that the $A could be stabilising soon – but mainly it’s a reflection of the international quality of our banks’ earnings.

I’ve stopped counting the number of times I’ve heard the story about Australia’s banks not having much of an outlook.

The banks power along regardless while the fundamentals of the economy remain reasonably sound. Looking good

As the Reserve Bank’s financial system review made clear, we’re actually looking pretty good.

(There is a sub-set to the banking analysts’ game: with just four major subjects, you go over-weight on two of them, wait for them to rise a bit, then switch your recommendation to the other two as representing under-priced value. It seems you can go on doing this for many years.)

Meanwhile, back at the August retail sales numbers that became an excuse for a bit of a kerfuffling because the ABS seasonally adjusted result came in lower than “expectations” (0.1 per cent growth compared with a prediction of 0.4 per cent growth), there’s more than a little nonsense involved in the building of those “expectations” in the first place and the continued emphasis on seasonally adjusted numbers that the market knows are, well, variable.

The Bloomberg economists’ survey of “expectations” generally includes a bunch of people taking a wild guess and four big banks who really should know the answer before the ABS.

The CBA, ANZ, NAB and Westpac have spent untold fortunes building flash computer systems that are supposed to be capable of making use of the “big data” that the banks capture. Five possibilities

Each of the Big Four has a large enough slice of the market to know what consumers are really up to – if their glorified abacuses are performing as they should.

Thus there are five possibilities: the banks’ CIOs haven’t delivered the systems they are forever promising;the CIOs don’t trust the economists with the numbers;the bank economists all play dumb in their forecasts so as not to give away perceived internal big data advantages;the ABS seasonally adjusted series is rather random and should not be taken too seriously; ora bit of all of the above.

Anyone remember the last couple of monthly labour market figures? So why would you think seasonally adjusted retail sales would be any more reliable?

There was nothing surprising about the trend series retail sales numbers – a gain of 0.2 per cent following a gain of 0.2 per cent.

On the ABS “Key Features” page the trend figures are printed ahead of the seasonally adjusted and there’s a graph of the trend series but not of the seasonally adjusted. Maybe the ABS is trying to tell us something.

And as for what the retailers are quietly saying, the minutes of the Reserve Bank September 2 board meeting included this line:

“Information from liaison with firms suggested that the value of retail sales had increased over July and August.”

That ABS trend series graph shows retails sales growth improved from a weak April and May to be steady at 0.2 per cent for the next three months.

This story Administrator ready to work first appeared on Nanjing Night Net.