18/07/2011: Some thoughts on Irish stocks bubble

There is a classic defined relationship between the various stages of bubble formation and markets responses, as illustrated in the chart from (source here) below.

Of course, there is an argument to be made that ‘normal’ bubbles are driven by either information asymmetries or behavioural ‘exuberance’ or both, and are, therefore, significant but temporary departures from the steady state ‘mean’ growth trend. The return to the mean, thus implies the end of the correction phase, as also shown in the chart below.


Of course, one can make an argument that what we have experienced in the case of Ireland is more than a simple bubble, but a structural break underwritten by underlying fundamentals, such as lower permanent rate of growth.

Irish GDP grew 8.82% cumulative in the period 2003-2010 in terms of constant prices or annualized rate of growth of 1.215%. In per capita terms current prices it grew by 14.85% cumulatively and at an annualized rate of 1.998%. Taken from these rates, from 2003 on through today, the average expected value of IFIN should be around 8,898 (mid-point between 8,659 and 9,139 implied by above rates from the ‘Smart Money’ period mid-point valuation). Note that, crucially, the new mean post-bubble bursting should be at least at or above the ‘Smart Money’ end-of-period valuations.

This is certainly not the case with Irish financials as shown in figure below:
Note that three forecasts (my own calculations, so treat as indicative, rather than absolute) provided assume that the average annual growth rate of 1.998% (upper forecast from the starting point at 2003-2004 average), mean forecast (based on 1.215% annualized average growth, starting from 2003-2004 average) and lower forecast (based on 1.215% annualized growth, starting from 2000-2003 average). All three are well above the post-Despair peak.

What about other signs of a classic bubble?
In the run up to the Public Money phase, it is clear that IFIN shows a number of sell-offs and shallow bear traps, but these can be linked to higher overall volatility of the index.

For any period we can take, IFIN exhibits more volatility than either S&P or FTSE bank shares sub-indices. Historically, across indices (to assure comparable scale), IFIN standard deviation stands at 65.40 against S&P’s BIX at 36.84 and FTSE A350 Banks at 32.70. January 2003 through June 2006, IFIN standard deviation was 25.16 against that for BIX of 10.29 and FTSE A350B at 12.07. For the run up to the crisis period between June 2006 and June 2007, IFIN standard deviation was 15.66 against S&P’s BIX of 4.64 and FTSE A350B of 5.22. Lastly, during the crisis – from July 2007 through today, IFIN standard deviation was 56.40 against 28.07 for S&P BIX and 27.83 for FTSE A350B.

To see the relationship, or the lack there of between the volatilities, consider the following chart.
Even from the simple consideration of the rates of change, week on week, IFIN has the lowest correlation with the S&P Banking BIX index – with relatively low explanatory power. Things are even worse if we are to look at the downside risks. Chart below plots downside weekly movements for the three indices that correspond to market declines of 2% or more week-on-week. Again, you can see that both before and during the crisis, there is little relationship between downside risk to Irish financials and to S&P measure.
And the same story is formally confirmed by the Chart below which plots the pair-wise relationships between S&P BIX and FTSE A350 and IFIN.
So overall, IFIN data strongly suggests that we are not in a “normal” financial bubble scenario.

But what about that claim that Lehman's Bros collapse had influence on our banks shares? Recall, Lehman was in trouble since Spring 2008 and went to the wall on September 15, 2008. Also recall that the issues started with Bear Sterns troubles in March 2008 and JPMorgan Chase completed its acquisition of Bear Stearns on May 30, 2008. So let's take the data subset on extreme downward volatility for the period from May 2008 through September 2009. If Lehmans and/or Bear had much of an effect on Irish financials we should expect either one of the following two or both to hold:

1. Correlation between IFIN and S&P BIX to be large and significant
2. Correlation between IFIN and BIX to be larger in the period considered than over the history from 2003 through today.

Overall, evidence suggests that actually the opposite of both (1) and (2) above holds. In fact, based on data for weekly market declines greater than 2% (relatively significant events, but not really too dramatic by far), the period between Bear & Lehman collapse and the next 12 months, Irish financials were less impacted by the US financial shares movements than in the period of 2003-present overall. The impact of Lehmans & Bear on UK financials was stronger, although not dramatically strong, however.