Hi Denk, Thats interesting analysis. I must say that I prefer my...

Hi Denk,

Thats interesting analysis.  I must say that I prefer my method of using the “cases completed in the current year” as it makes it easier to see exactly the margins with the full benefit of hindsight.

I think the error in your analysis is that you are calculating and applying a “maximum case multiple” (10.1%) to a “point in time” expected completion case loan.  To compare like for like you would need to apply the “maximum case multiple” to the “maximum case expected completions”.   

For example if we take as the 10.1% “maximum case multiple” you should apply that to the “maximum case expected completions” i.e. 2017 ($1,077.9) 2018 ($2,221.8) 2019 ($1,508.7????) Total = $4808.4.  $4808.4 x 10.1% = $485.6.  While you say this is the maximum you can get to I think its probably less than the minimum of what it should be.

Whereas if you use the “% of resolved cases” method (15%) you would apply that to the “point in time” case expectation.  3,370.3 x 15% = $505.50.

In any event I think $500m is the ballpark we are looking for and both types of analysis tends to come up with roughly that number.

I think this “maximum approach” is problematic for a number of reasons:

1. How do you calculate the 2019 “maximum case expected” now? You can’t because that number is likely to go up next quarter. And while that may be right remember what we are trying to achieve from this analysis.  To value the book at this specific point in time.  e what is likely to come back in litigation income from our $3,370.3 Estimated portfolio value.
2. I think this type of analysis has a lot more overlap than the 15% “cases resolved” method because in any quarter if management believe a case is going to drag on to another year they simply move it out another year in these announcements. Having said that there is bias in the “cases resolved” method as well. The bias in the “cases resolved method” is that the Claim Value is the Claim value at June 30 the previous year.  As you can see with the ANZ case the claim value was already written down to $20m the previous year before the case was lost.  Im sure at one point they would have had a claim value well over $100m given that they lost $15.4m on the case.

Back-testing:

From your analysis what it does allow me to do is to backtest my 15% assumption against historical Expected Completion announcements and compare the results to the actual.

i.e if we were sitting here in 2011 I would be applying the 15% to the “point in time” 31/12/2010 case book and saying I expect 1614.5 x 15% = $242.17m to drop through to “litigation income” through 2011, 2012, 2013.  In actual fact $219.6m was the result.

Date of Book   Total EPV Expected (15%) completions   Total Actual completions   
31/12/10 $1,614.5m    $242m $219.6m
31/12/11 $1,616m   $242m $237.6m
31/12/12 $1,473m   $220m $212.1m
31/12/13 $1,942m   $291m $268m

When you consider we are looking at a period when we lost 2 of our biggest claim values over this period (Bank of QLD) completed 2014 and Bank fees (Claim value $140m) in 2016 I think the 15% method holds for point in time analysis.

Denk how do you download these images to Hotcopper?  I’ve got sensitivity analysis that I would share except I don’t know how to without the format of excel being warped.