ANZ.NZX - Anyone holding?

[Snoopy]

I now want to use the actual dividend data for FY2012 to FY2016 inclusive to build a 'scenario analysis' looking forwards.

The basis for my model is that market conditions over the last five years are broadly representative of what we might expect over the next five years. Yet in recent times, the capital base of the company has been expanded, because of Basel 3 international banking standards requiring a greater capital base to support the existing bank loan book. The expansion of ANZ's capital base has been as a result of a share issue, including a share purchase plan offer to existing shareholders and institutional share placement. Unfortunately for existing shareholders, this means that broadly the same income stream must now be distributed among the greater number of shares now on issue. In other words, unless there is a corresponding growth in profitability from the new share capital (there won't be, because the new capital is required to be used as a beefed up safety net), earnings per share can be expected to decrease in the future.

The following table shows what would have happened if the number of shares on issue today was constant over the previous five years.

	FY2012	FY2013	FY2014	FY2015	FY2016
Gross Annual Dividend (NZ Perspective) 'cps' (final) + (interim)	$1.42	$1.60.6	$1.93	$2.01.9	$1.95
Actual Number of Shares on Issue EOFY	2,744m	2,759m	2,757m	2,903m	2,927m
Actual Dividend %ge of 'Cash Profit'	69.3%	71.4%	67.4%	68.6%	81.9%
Scenario Number of Shares on Issue EOFY	2,927m	2,927m	2,927m	2,927m	2,927m
Scenario Adjusted Gross Annual Dividend (NZ Perspective *) 'cps' (final) + (interim)	$1.33.1	$1.51.4	$1.81.8	$2.00.2	$1.95

There is a problem with the table above. We have been told for a fact that to continue to meet 'modern capital requirements', the dividend will be reduced to a payout ratio of 60 to 65% of the declared 'cash profit' The table tells us that this percentage has been exceeded for all years over the past five years. This, in all likelihood, means that using the cash dividends listed above as indicative for the future means that we are going to forecast future cash dividends to be too high. So what happens if we reduce the payout ratio on those historical records down to 62.5% (the mid point of the 60% to 65% future declared dividend target range) of 'cash profit'?

* In this instance "NZ Perspective" means dividends continue to be expressed in $A, but the $A dividend is a 'gross dividend' because NZ does not recognize Australian Franking Credits. In addition NZ imputation credits are added to the $A dividend using the exchange rate $NZ1 = $A0.95.

SNOOPY

[macduffy]

[QUOTEThe basis for my model is that market conditions over the last five years are broadly representative of what we might expect over the next five years.][/QUOTE]

If only! A nice assumption but I'm not sure that "retrospective" forecasting has much value.

[Snoopy]

Snoopy"The basis for my model is that market conditions over the last five years are broadly representative of what we might expect over the next five years."

If only! A nice assumption but I'm not sure that "retrospective" forecasting has much value.

I wrote that because I wanted readers to be aware of the limitations of the technique. Despite the obvious limitation that you cannot say that any one year will be like any other year, it becomes more useful when you look at historical averages over many years (for this is what I am doing). In a broadly static market I like to use 'actual data' not 'extrapolated data'. My philosophy in doing this is to recognise that while this is not a great technique to use, it still compares well when put up against every other technique available!

SNOOPY

[percy]

I wrote that because I wanted readers to be aware of the limitations of the technique. Despite the obvious limitation that you cannot say that any one year will be like any other year, it becomes more useful when you look at historical averages over many years (for this is what I am doing). In a broadly static market I like to use 'actual data' not 'extrapolated data'. My philosophy in doing this is to recognise that while this is not a great technique to use, it still compares well when put up against every other technique available!

SNOOPY

Yeah right.!>>>lol.

[percy]

[QUOTEThe basis for my model is that market conditions over the last five years are broadly representative of what we might expect over the next five years.]

If only! A nice assumption but I'm not sure that "retrospective" forecasting has much value.

[/QUOTE]

Useful, but a hindreance as markets are forward looking.
Foresight is more profitable than hindsight.!

[Snoopy]

Foresight is more profitable than hindsight.!

Yes, but in a broadly stable market, how do you create your foresight? I look to the past for future inspiration, while keeping in mind that going too far back into the past is also getting distant from today's (and tomorrow's) reality. Five years is about the shortest company history I am comfortable with if I am looking for a track record. With some companies, I look at records going ten years back. With ANZ, given the Basel 3 capitalisation requirements in force today, my judgement is that going back any further than five years in the record might not be representative of 'the modern banking environment'.

Some brokers just take last year's result for any company and add a fudge factor and call that a forecast.

If you have a better 'forward looking system', then let's hear what it is!

SNOOPY

[Snoopy]

So what happens if we reduce the payout ratio on those historical records down to 62.5% (the mid point of the 60% to 65% future declared dividend target range) of 'cash profit'?

Note:

1/ All dollar figures listed are in Australian dollars, unless specified otherwise.
2/ Because the New Zealand Dividends are not fully imputed, I am assuming the the gross value of imputation credits passed on will not be affected as the gross dividend is reduced.

	Scenario FY2012	Scenario FY2013	Scenario FY2014	Scenario FY2015	Scenario FY2016	Average Across Five Scenarios
Annual Declared Dividend 'cps' [(final) + (interim)] (imputation credits excluded)	$1.42	$1.52	$1.74	$1.81	$1.75
Actual Dividend %ge of 'Cash Profit'	69.3%	71.4%	67.4%	68.6%	81.9%
Scenario First Adjustment: Declared Annual Dividend 'cps' [(final) + (interim)] (62.5% payout ratio) (A)	$1.28	$1.33	$1.61	$1.65	$1.34
Actual Number of Shares on Issue EOFY (B)	2,744m	2,757m	2,757m	2,903m	2,927m
Scenario Number of Shares on Issue EOFY (C)	2,927m	2,927m	2,927m	2,927m	2,927m
Scenario Second Adjustment: Normalise Dividend for Today's Number of shares (A) x [(B)/(C)] => {D}	$1.20	$1.25	$1.52	$1.64	$1.34
Actual Imputation Credits (Conversion $NZ1- =$A0.95) (E)	0	$(0.09 x 0.95)	$(0.20 x 0.95)	$(0.22 x 0.95)	$(0.21 x 0.95)
Scenario Third Adjustment: Normalise Imputation Credits for Today's Number of shares (E) x [(B)/(C)] => {F}	0	$0.08	$0.18	$0.21	$0.20
Gross Dividend: NZ Investor Perspective {D} + {F}	$1.20	$1.33	$1.70	$1.85	$1.54	$1.524

SNOOPY

[macduffy]

Yes, but in a broadly stable market, how do you create your foresight? I look to the past for future inspiration, while keeping in mind that going too far back into the past is also getting distant from today's (and tomorrow's) reality. Five years is about the shortest company history I am comfortable with if I am looking for a track record. With some companies, I look at records going ten years back. With ANZ, given the Basel 3 capitalisation requirements in force today, my judgement is that going back any further than five years in the record might not be representative of 'the modern banking environment'.

Some brokers just take last year's result for any company and add a fudge factor and call that a forecast.

If you have a better 'forward looking system', then let's hear what it is!

SNOOPY

No better system but I regard it as a waste of effort to try to forecast five years ahead. Incidentally, are you forecasting that Basel 3 is the end of the story regarding capital requirements for banks and that it will be in force, as envisaged today, five years hence, ie extrapolated 5 years out?

[Snoopy]

No better system but I regard it as a waste of effort to try to forecast five years ahead.

I think you are misinterpreting what my objective is. I am using five years of historical data to forecast one year ahead.

I have labelled five scenarios as Scenario FY2012, Scenario FY2013, Scenario FY2014, Scenario FY2015 and Scenario FY2016. The underlying business metrics that generated this data came out in one year increments over five years of chronological order. However, as future scenarios, this historical order no longer has meaning. I am assuming that each scenario has the same liklihood of occurring as any other scenario. And to take this into account I am going to consider an 'annual average figure' to make my future estimates of value.

Of course since I am assuming a 'broadly static market', you could say that my prediction for five years ahead is the same as that for one year ahead, and in this sense your original assertion is correct. But as new data comes along in the interim years, that prediction for 'that year that is five years ahead' now will of course be revised.

Incidentally, are you forecasting that Basel 3 is the end of the story regarding capital requirements for banks and that it will be in force, as envisaged today, five years hence, ie extrapolated 5 years out?

I am assuming the effect of Basel 3 will be constant into the forseeable future, yes. You have to take that into consideration when assessing any predictions from my modelling.

SNOOPY

[Snow Leopard]

So when do we actually get the forecast of future earnings, dividend payout and do we get a valuation with that?

I can hardly handle the suspense any longer.



Best Wishes
Paper Tiger