Author Topic: Dividends vs Capital Gains once again  (Read 3254 times)

Ursus Major

  • Stubble
  • **
  • Posts: 157
  • Location: Silicon Valley, CA
Dividends vs Capital Gains once again
« on: January 24, 2016, 03:21:00 PM »
Hi all,

I'm a new member of this forum and recently I was pointed to an old thread about comparing receiving dividends vs. selling stocks for capital gains: http://forum.mrmoneymustache.com/investor-alley/ive-decided-on-vanguard-but-need-some-help-please/msg584803/#msg584803. It compares the results of two fictitious companies, Divcorp and Gapgains, of which the former pays dividends and the latter does not.

Since the thread is quite old, I've decided to start a new thread on this topic. I'm sorry, if I'm beating a dead horse here, but I don't quite follow the conclusion in the thread above and like to have this resolved for me as well.

The central claim of the post above (analyzing the consequences of an adverse event and the associated fall in earnings and stock price) appears to be the following:

Of course, this doesn't make Capgains "better" than Divcorp, because Divcorp provided you with retirement income via its dividend. But when you are forced to "eat into principal" of your Capgains holding to fund your retirement, you do so from a position where Capgains has fallen less-far than Divcorp. "Eating into Capgains' principal" at that point only makes your holding shrink to be as small as Divcorp, not smaller. Your eaten-away holdings in both companies will have an equally hard time recovering from those losses; neither is in a stronger position than the other.

However when I run the number, I see a different result. Yes, the stock price of Divcorp will fall further than Capgains, because Capgains retains earnings. However that doesn't help the investor, because she has to sell shares of Capgains at a low price and will be worse off in the long run.

I've tied to follow the example in the post above as close as possible and use the following assumptions:
- The investor holds 10% of each company at the start of this period (end of 2016). They will receive the 5% dividend from Divcorp, and will sell an equivalent amount of shares from Capgains.

- Divcorp is just starting to pay that dividend, so their balance sheet (including retained earnings) is identical to Capgains.

- Investors expect a 6% earnings yield and will pay 1x book value for the balance sheet, which is assets-debts and includes retained earnings. (I am excluding working capital here, so my balance sheet only lists "excess" assets that are not needed for the business. The result is the same with working capital included, since the investor would only pay something for excess capital). I am booking the earnings inflow to the books on Jan 1 of the following year (to stay close to the original assumption that the stock market value of each company is $1M).

Here is my scenario:
1. Original starting point: Bad news in late Dec 2016: 2016 earnings are unaffected, but 2017 and forward earnings will fall by 50% for both Divcorp and Capgains to $30k. In an slight modification I gave both companies an extra $50k in assets ($5k) per share. That brings the stock market value to $1.05M for each company.

2. 2017 and 2018 play out as expected, but at the of 2018 Cleveland has been rebuild, the companies gizmos are gaining in popularity and 2019 earnings are on the rebound and are expected to be back to $60k a year for both companies.

3. The investor would like to withdraw $5000 every year from Divcorp and and the same amount from Capgains, either from dividends or from stock sales (or both, if necessary). Since Divcorp pays out the dividend on Jan 15th, she picks that day for the withdrawal from both companies.
4. I expect DivCorp to maintain its dividend and pay it out of existing assets.

Please see the attached XLS for the calculation. It is modular enough to allow you to change several of the assumptions.

When I run the numbers, I see that the investor in Divcorp is already coming out ahead on the dividend payment on 1/15/2018 and will pull especially ahead, if the earnings recover to their previous value.

The only way that Capgains will match Divcorp on 1/15/2018 is, if I assume that both companies will earth the same yield on their balance sheet assets as the investors expect as earnings yield, that is 6% (Field D3 in the XLS). However once the earnings bounce back, then the investor in Capgains does not equally participate in the corresponding price bounce back, since he has fewer shares.

Perhaps there is a bug in my XLS. If so, please tell me where, as I'd like to resolve this for myself.

Interest Compound

  • Pencil Stache
  • ****
  • Posts: 658
Re: Dividends vs Capital Gains once again
« Reply #1 on: January 24, 2016, 03:35:01 PM »
From cells H12 to H13, Divcorp price increases by 1.9259x

From cells H27 to H28, Capgains price increases only by 1.7812x

If the companies are both identical, and both became equally more valuable, price should increase by the same ratio, or percentage. Instead you increased them by the same dollar amount.

Ursus Major

  • Stubble
  • **
  • Posts: 157
  • Location: Silicon Valley, CA
Re: Dividends vs Capital Gains once again
« Reply #2 on: January 24, 2016, 03:50:54 PM »
From cells H12 to H13, Divcorp price increases by 1.9259x

From cells H27 to H28, Capgains price increases only by 1.7812x

If the companies are both identical, and both became equally more valuable, price should increase by the same ratio, or percentage. Instead you increased them by the same dollar amount.

My assumptions are that investors demand a 6% earnings yield and will pay 1 times book value for net excess assets (or subtract the same for debt) on the balance sheet. From H12 to H13 the price increases by $50 per share, because for both companies earnings are going up the same amount ($3 per share). Investors are not going to pay more for excess assets on the balance sheet, just because earnings are going up.

So I believe this correctly reflects the assumptions that I've made (and that was made in the original post that I am quoting).

If you think these assumptions are wrong, then you would need to explain, why a (supposedly) rational investor would all the sudden start paying more for excess assets on the balance sheet.

Interest Compound

  • Pencil Stache
  • ****
  • Posts: 658
Re: Dividends vs Capital Gains once again
« Reply #3 on: January 24, 2016, 04:30:41 PM »
From cells H12 to H13, Divcorp price increases by 1.9259x

From cells H27 to H28, Capgains price increases only by 1.7812x

If the companies are both identical, and both became equally more valuable, price should increase by the same ratio, or percentage. Instead you increased them by the same dollar amount.

My assumptions are that investors demand a 6% earnings yield and will pay 1 times book value for net excess assets (or subtract the same for debt) on the balance sheet. From H12 to H13 the price increases by $50 per share, because for both companies earnings are going up the same amount ($3 per share). Investors are not going to pay more for excess assets on the balance sheet, just because earnings are going up.

So I believe this correctly reflects the assumptions that I've made (and that was made in the original post that I am quoting).

If you think these assumptions are wrong, then you would need to explain, why a (supposedly) rational investor would all the sudden start paying more for excess assets on the balance sheet.

Seems like you're looking to be argumentative. This will be my last post in this thread :)

When price went down from 105, both companies became less valuable by the same percentage (42%). When the two companies became more valuable, they didn't increase by the same percentage. Check your math. I suspect F27 is where you made your mistake.

Good luck!

seattlecyclone

  • Walrus Stache
  • *******
  • Posts: 5130
  • Age: 35
  • Location: Seattle, WA
    • My blog
Re: Dividends vs Capital Gains once again
« Reply #4 on: January 24, 2016, 05:08:54 PM »
I think the general assumption is that companies that retain earnings will use them to invest back into growing the business, and won't just leave the money as cash in the bank. Therefore you should assume that the company retaining its earnings will increase earnings by a larger dollar amount than the company paying dividends. If the company can't reinvest those earnings to make earnings grow even $1 faster than the company that paid dividends, then they should have just gone ahead and paid dividends!

neil

  • Stubble
  • **
  • Posts: 216
Re: Dividends vs Capital Gains once again
« Reply #5 on: January 24, 2016, 05:39:22 PM »
Both companies start off equivalent.  From an earnings perspective both companies appear to have the same behavior.  However, the profitability of Divcorp is stable while Capgains is shrinking when looking at metrics like ROE/ROA.

The reason retained earnings are theoretically equivalent only holds up because the company invests that cash flow back into the company to create earnings growth.  Not knowing why Capgains is not growing earnings, they are either letting cash depreciate in the bank, needed to invest money to retain existing business or failed at whatever new investments were made.  In any of these scenarios, Capgains is destroying shareholder value.

If you think that shouldn't matter in your hypothetical test case, consider there are two moving parts to Capgains - core business X that has solid business earnings and side business Y that is doing nothing except accumulating cash.  Even though X is fine and Y is growing, every time you sell you are giving up part of your claim on X and Y and the earnings you need for expenses are only available to the X part of the business.  The investor wants 6% return, but he is not getting it because only X is returning 6% and he is slowly selling off parts of X and replacing with Y which has zero return.  His portfolio will eventually go to zero for that reason.  The Divcorp investor is only "selling" Y because the company is spinning that off to the investor directly rather than retaining it and doing nothing - and Divcorp investor holds onto his original holding of X the entire time.

Ursus Major

  • Stubble
  • **
  • Posts: 157
  • Location: Silicon Valley, CA
Re: Dividends vs Capital Gains once again
« Reply #6 on: January 24, 2016, 07:31:19 PM »
Seems like you're looking to be argumentative. This will be my last post in this thread :)

When price went down from 105, both companies became less valuable by the same percentage (42%). When the two companies became more valuable, they didn't increase by the same percentage. Check your math. I suspect F27 is where you made your mistake.

It seems that I upset your with my reply to your post. I'm sorry that this happened. All I was trying to say is that I made the calculations I did for a particular reason and I was trying to convey that logic behind it.  I am surprised at these results myself and I am looking forward to a case as to where I am going wrong.

When the price went down from $105, it went down to $56 for Divcorp for a loss of 46.7% and to $61 for Capgains for a loss of 41.9%. So Divcorp actually became less valuable than Capgains as a company (due to less retained earnings).

F27 is actually pretty simple, it takes the previous net assets and add earnings and interest (if any) and subtracts paid out dividends. So $11 net assets from 12/31/2017 plus $3 in net earnings for 2017 (plus zero interest minus zero dividend payout) makes $14.

Ursus Major

  • Stubble
  • **
  • Posts: 157
  • Location: Silicon Valley, CA
Re: Dividends vs Capital Gains once again
« Reply #7 on: January 24, 2016, 08:01:55 PM »
I think the general assumption is that companies that retain earnings will use them to invest back into growing the business, and won't just leave the money as cash in the bank. Therefore you should assume that the company retaining its earnings will increase earnings by a larger dollar amount than the company paying dividends. If the company can't reinvest those earnings to make earnings grow even $1 faster than the company that paid dividends, then they should have just gone ahead and paid dividends!

This is an excellent point and I observe its effect, when I set the interest rate (for excess net assets) to 6% (Cell D3). Then the investor will also have received the same result on 1/15/2018. But it still is unequal on 12/31/2019, after earnings expectations have been reset.

Both companies start off equivalent.  From an earnings perspective both companies appear to have the same behavior.  However, the profitability of Divcorp is stable while Capgains is shrinking when looking at metrics like ROE/ROA.

The reason retained earnings are theoretically equivalent only holds up because the company invests that cash flow back into the company to create earnings growth.  Not knowing why Capgains is not growing earnings, they are either letting cash depreciate in the bank, needed to invest money to retain existing business or failed at whatever new investments were made.  In any of these scenarios, Capgains is destroying shareholder value.

If you think that shouldn't matter in your hypothetical test case, consider there are two moving parts to Capgains - core business X that has solid business earnings and side business Y that is doing nothing except accumulating cash.  Even though X is fine and Y is growing, every time you sell you are giving up part of your claim on X and Y and the earnings you need for expenses are only available to the X part of the business.  The investor wants 6% return, but he is not getting it because only X is returning 6% and he is slowly selling off parts of X and replacing with Y which has zero return.  His portfolio will eventually go to zero for that reason.  The Divcorp investor is only "selling" Y because the company is spinning that off to the investor directly rather than retaining it and doing nothing - and Divcorp investor holds onto his original holding of X the entire time.

I fully agree with you and trying to rack my brains on how to model this scenario appropriately. I prepared for that by adding an interest/debt rate and (see my comment above) and (as mentioned above) at a 6% rate the results are equal before the earnings rebound.

In order to be equal on the last day (12/31/2020) the Interest Rate on Assets has to be around 21.6%. Still trying to get a mental model as to why...

Some random thoughts that need further investigation:
  • a) At the moment I have not explicitly modeled any ROA/ROE into the model, but are there already some implicit assumptions about ROA/ROE in the model?
  • b) In order to have constant ROA (or ROE) both companies need to shet assets quickly on a major earnings drop (and build them up again just as quickly on an earnings rebound). Can (and will) a company just drop 50% of their assets and ramp them up again by 100% on short notice? Or will they (like both Divcorp and Capgains do) wait it out, if it's a short-term solution.
  • c) How do I need to model this scenario that the results for Divcorp and Capgains become the same again on 12/31/2020? And then would this model reflect reality (see point a) )?
Looking forward to any suggestions.

faramund

  • Bristles
  • ***
  • Posts: 331
Re: Dividends vs Capital Gains once again
« Reply #8 on: January 24, 2016, 10:06:57 PM »
I tend to think, in theory it doesn't matter. But in practice, I think its significant, as a company that says it is profitable and consistently pays dividends has a higher probability of being profitable, then another company that also says its profitable, but doesn't.

So I also flip this, and say, well, if it doesn't matter whether you go for a dividend or non-dividend stock, I might as well buy the dividend stock for peace of mind.

There is also statistical support for dividends, the first is O'Shaughnessy's "What works on Wall Street",
where he essentially tries out different stock-market strategies (its about my favorite stock market book)

For dividends, he divides all stocks into 10 deciles. The average return of all stocks, 1927-2009, he found to be 10.5%. The top percentile of dividend stocks returned 11.8%, the lowest (smallest/no dividend) returned 9.7%.

In a similar vein, in Australia, there is VAS (all Australian stocks), and VHY (just high yield stocks). Since VHY was created, it outperforms VAS (although over the last 2 years - that's reversed)