Wednesday 9 February 2011

Return on Equity: Errrm? Help!

I do not profess to be an academic or a professional analyst, but I like to understand the numbers and ratios that I am looking at in terms of what they really mean and whether they can be misinterpreted or manipulated (by others!).

Received wisdom seems to have Return on Equity (RoE) as an accepted measure of how good a company is at delivering profits for shareholders, which seems fair enough.

In essence, it is calculated by dividing Net Income (Profit After Tax) by Net Assets (Equity funds (attributable to ordinary shares)) on the opening balance sheet. So, if you started with Net Assets of 100 and generated Net Income of 15, your RoE is 15% for the year (one of my Rules). (Note: I use the opening balance of Net Assets, but I have seen others who take an average of the opening and closing Net Asset position). 

Let's consider a worked example and I make no apologies for trying to keep it as simple as possible:

  • a company generates £10m of PAT each year for three years;
  • it has opening Net Assets (ordinary shareholders' equity; non-cash) of £50m; and
  • the Market values the company on a P/E of 12x PAT.

A Basic analysis looks like this:

BASIC
Opening
Yr 1
Yr 2
Yr 3
ROE


"Value"

PAT (£'m)

10
10
10



120

DIVIDENDS (£'m)

0
0
0




NET ASSETS bal b/f

50
60
70
NET ASSETS bal c/f
50
60
70
80



30
Ave
ROE

20%
17%
14%
17%


150



This assumes that each year's profits are retained by the company. The RoE declines each year as the company is not "sweating" its assets hard enough due to the growing surplus cash pile.

From a valuation perspective, the company would be valued at £150m, assuming earnings are valued at £120m (12x PAT) and surplus cash (£30m) is valued at par.

However, let's throw in some dividends. Assuming that all post-tax profits are distributed as dividends each year, the analysis would look something like this: 

DIVIDENDS
Opening
Yr 1
Yr 2
Yr 3
ROE


"Value"
PAT (£'m)

10
10
10



120
DIVIDENDS (£'m)

10
10
10



30
NET ASSETS bal b/f
50
50
50
NET ASSETS bal c/f
50
50
50
50
Ave
ROE

20%
20%
20%
20%


150



The RoE remains steady each year at 20% as surplus cash is not allowed to build up. From a value perspective, the company would be valued at £120m and the shareholders would have received dividends of £30m, resulting in "value" of £150m (ignoring time value of money and personal tax implications). Same "value", better RoE.

Now let's throw some share buy-backs into the mix. Assuming the company spends all of its profits on share buy-backs, the analysis would look something like this.

BUY-BACKS
Opening
Yr 1
Yr 2
Yr 3
ROE


"Value"
PAT (£'m)

10
10
10



152
BUY-BACKS (£'m)

10
10
10
NET ASSETS bal b/f

50
46
42
NET ASSETS bal c/f

50
46
42
39
Ave
ROE

20%
22%
24%
22%


152















The RoE goes up each year as the net assets figure shrinks due to the buy-backs. Interestingly, the value increases slightly (could be my maths) even despite the fact that the company is essentially wasting shareholders' money by buying back expensive shares in the open market - ie £10m of year 1 share purchases only reduces the Net Assets by £4m (assuming it would not otherwise be held as surplus cash). However, for each £1 of share capital bought back, there is a multiplier effect on value as EPS goes up (due to profits being spread over fewer shares). 

So, which is best? From a big-picture perspective, the same company is producing the same level of profits from what is essentially the same operating capital base, so there should be little difference. From a RoE perspective and "value creation" perspective, scenario 3 appears to be superior with an average RoE 30% higher than the basic scenario, although if I came across this scenario, I would be put off by the market value being at a high premium to net assets and the lack of a dividend! Maybe screening for a high RoE could hide suitable opportunities?

There is also the small issue of leverage (debt) which helps to improve returns to shareholders through the introduction of lower cost funds (compared to equity).  There is probably an optimal balance between debt and equity for each company to maximise the impact on its RoE.

Conclusion

  • RoE is an important, but crude measure of how much profit is being generated for shareholders;
  • A rising RoE does not mean that a company is generating higher levels of profits per se and vice versa;
  • ROE can be distorted by dividends, share buy-backs and leverage - for instance, see GlaxoSmithKline which has an unfeasibly high RoE due to leverage and all net profits being distributed through dividends and share buy-backs;
  • There may be opportunities in finding companies with surplus cash on the balance sheet and an accelerated buy-back policy - an increasing RoE might capture the attention of more investors and drive up the price;
  • Return on Capital Employed (ROCE) might be a better measure of how much return a company generates through employing debt and equity. (EBIT/(Debt+Equity). I will research some more and see whether that is more appropriate for my Rules; and
  • I could be barking up the wrong tree!
Any thoughts or comments are most welcome

9 comments:

  1. A few things to think about:
    1. ROE is the most basic measure of the return the shareholders receive on the cash they have invested in the business (either via the intial equity raising, or because the company has decided to retain its profits).
    2. From the company's perspective, there is no difference between paying dividends and making share buy-backs (though there may be differences in the tax treatment faced by the shareholders).
    3. I believe your buy-back scenario is incorrect in assuming that the equity base shrinks over time. This would only happen if the buyback was greater than that year's net profit. If the buy-back is equal in sized to the annual profit, then net assets (book value of equity) would remain the same. However, buy-backs do impact the market value of equity in that assuming a constant P/E ratio, the reduced share count would mean a higher price for each share.
    4. An important relationship involving the ROE is the sustainable growth rate. G = ROE*RR where g = sustainable growth rate; RR = retention ratio = 1 - dividend pay-out ratio. What this equation means is that the company can grow by retaining its earnings within the business and reinvesting them in projects that will achieve the prevailing level of ROE. The more cash retained and higher the level of ROE, the faster the potential growth rate of earnings.
    5. I wouldn't describe an ROE as being "distorted" by dividends, share buy-backs and leverage. I would view company's that maximised their ROE by minimising their use of equity capital in a positive light. There's nothing I hate more than a company with a huge net cash balance and management refuse to do anything with it - it is essentially a negative carry trade, acting as a drag on shareholder returns.
    6. I'd recommend looking at the DuPont model, which breaks-down the ROE into its individual components. http://www.investopedia.com/articles/fundamental-analysis/08/dupont-analysis.asp

    Hope that helps?

    ReplyDelete
  2. Cautiousbull

    Many thanks for such comprehensive thoughts. In terms of some of the points you have raised:

    3 - I had assumed that surplus cash would be distributed or used for buy-backs and therefore would not be included in net assets to start with. To the extent that it is used for buy-backs, the £10m would be spent acquiring shares in the market. However, the nominal value of those shares would only be £4m given that the shares are trading on a premium to net assets. In this sense, the action is wasteful from a net asset value perspective as the company has acquired assets over and above the 'book cost'. One thing I'm not clear on, is what happens to the remaining £6m, is this adjusted through the SPA or other reserves?

    However, this is more than compensated by the reduction in the number of shares, which would drive the share price higher.

    4 + 6 - interesting things to consider. I haven't come across DuPont before, so I'll have a good look.

    5 - maybe "influenced" or "effected" might be more appropriate words! Maybe there is an opportunity to find cash rich, low ROE companies where there is a catalyst to utilise (profitably) the surplus cash?

    ReplyDelete
  3. To return to point 3, the business in question has retained earnings of nil, yet you show the book equity to be decreasing. Given double-entry book-keeping, one of two things must also be occurring: (1) Liabilities must be increasing; or (2) assets must be decreasing. Neither is apparent from this analysis. If a company has used all its annual net profit to acquire its own shares, then net assets will only decrease to the extent they have been written-down or sold below book-value during the year in question. This factor would be the same regardless as to whether or not: (1) the company paid dividends or repurchased shares; and (2) the company repurchased its stock above or below book value.

    The Balance Sheet effects of a £10m dividend and a £10m share buy-back are the same: Cash (Current Assets) -£10m; Equity (Retained Earnings) -£10m. If the company has also earned £10m during the year, then this will replenish both assets and equity to their original levels.

    ReplyDelete
  4. Taking thing back a bit...why is the valuation 152 in example 3? Is it 120 + (42-10)? Please explain then I can read on...thanks

    ReplyDelete
  5. Hi V4V

    I think there is a mistake in my methodology, but in essence I was calculating it as: £10m PAT x 12x PER x a factor of 50/39 to reflect the fact that three years' worth of buy-backs would have reduced the share capital by about £11m in nominal value.

    For instance in year 1, £10m of PAT would be used to buy back shares in the market, which would be valuing the company at 12x PAT (£120m), the share purchase would acquire (£10/£120) c8.3% of the share capital (total nominal value of £50m, therefore c£4m of nominal value bought back and cancelled). Carry on for 3 years and you arrive at a total of c£11m in nominal value, hence the 39 (being 50-11ish).

    In reality my logic is flawed as I should have removed £10m from net assets (c£4m through share capital and c£6m through other reserves) in year 1, and repeated for years 2 and 3.

    Thanks

    ReplyDelete
  6. Hi Yorkiem. Ah, I see - thank you. I appreciate it was simplified. In reality, I assume in share buy backs that the earnings are 'swapped' for share capital (nominal & premium). As a result are you not restructuring the balance sheet with the net effect being the same net assets but with less shares outstanding? Your accounting expertise is appreciate here...thanks again

    ReplyDelete
  7. Hi V4V. My scenario was a kind of hybrid in that I had assumed the cash would be spent on dividends or buy-backs and was not treated as surplus cash on the balance sheet, which is why I only showed one half of the transaction.

    As cautiousbull rightly points out, if you use surplus cash to buy back shares, you are reducing the net assets by the amount of cash that you spent. This will increase the ROE next year if you have the same level of profits as you have a lower starting net asset position.

    The share price should increase as EPS will increase as a direct consequence of earnings being spread over fewer shares (although you have reduce net assets).

    You can see why companies like share buy-backs, particularly as lots of directors' performance targets are linked to share price performance. Buy-backs allows them to do this without increasing the real value of the company.

    ReplyDelete
  8. So good topic really i like any post talking about Business Ideas and Advices but i want to say thing to u Business not that only ... you can see in Business Business and Profit and more , you shall search in Google and Wikipedia about that .... thanks a gain ,,,

    ReplyDelete
  9. You also need to consider the use of dividends. I think this is a 'hole' that a lot of investors miss. To compare like with like surely you should assume that in the 'all dividend' scenario the investor is going to re-invest the dividends into shares.

    A lot of people get expected by the eps growth (or PEG) ratio and therefore put companies that are growing their eps on a premium. However as you have demonstrated the eps can grow not due to management skill but rather balance sheet 'optimisation'.

    Sharebuybacks were large in the tech boom. Another fallacy that a lot of investors and managements miss is that a share buyback only adds value if the shares are bought at a discount to intrinsic value. However if they are at a discount then why are shareholders not reinvesting their dividends in this? During the tech boom a lot of managements believed their own hype and bought shares at a large premium to intrinsic value.

    Intrinsic value is not always clear to calculate - and why should managements have the skill set or knowledge to compute a dispassionate robust and conservative intrinsic valuation.

    Roddy.

    ReplyDelete