Sure, there are fewer shares - but the value of the company has gone down (it has less free cash, i.e. fewer assets) and the two effects should exactly offset each other.
An example would be a stock you value at $10 per share. It has $5 per share of excess cash, but you attribute $3 of value to the cash. This is because you discount 20% for dividend taxes were it to be paid out, and a further 20% for risk it may never be dividended or that it might be invested poorly.
So if the company pays out the actual $5, netting you $4, the remaining shares should still be worth $7 in your valuation estimate, meaning $1 in value is created.
This is of course pedantic, because it can be argued that there should be no risk discount for cash, and if there is certainly not 20%. My argument would be I'd value $1 of cash in Buffett's hands as worth more than $1, and in the hands of many public CEOs hands at less than $1. There is too much self-dealing by public management, and they usually possess little investment expertise outside their own business.
Does Apple have any car development expertise or unique IP? I'd say the risk of them blowing $40B in that market is significant.
So I'd argue that a small risk discount exists in most cases and is recaptured upon dividend.
This McKinsey article is old but explores the topic in some detail.
http://www.mckinsey.com/business-functions/strategy-and-corp...
--
"Sending signals
The market responds to announcements of buybacks because they offer new information, often called a signal, about a company’s future and hence its share price.
One well-known positive signal in a buyback is that management seems to believe that the stock is undervalued. Executives can enhance this effect by personally purchasing significant numbers of shares, since market participants see them as de facto insiders with privileged information about future earnings and growth prospects. A second positive signal is management’s confidence that the company doesn’t need the cash to cover future commitments such as interest payments and capital expenditures.
But there is a third, negative, signal with a buyback: that the management team sees few investment opportunities ahead, suggesting to investors that they could do better by putting their money elsewhere. Some managers are reluctant to launch buyback programs for this reason, but the capital market’s mostly positive reaction to such announcements indicates that this signal isn’t an issue in most cases. In fact, the strength of the market’s reaction implies that shareholders often realize that a company has more cash than it can invest long before its management does.
Therefore, the overall positive response to a buyback may well result from investors being relieved that managers aren’t going to spend a company’s cash on inadvisable mergers and acquisitions or on projects with a negative net present value. In many cases, a company seems to be undervalued just before it announces a buyback, reflecting an uncertainty among investors about what management will do with excess funds.
Such shareholder skepticism would be well founded. In many industries, management teams have historically allocated cash reserves poorly. The oil industry since 1964 is one example (Exhibit 4): a huge price umbrella for much of this period, courtesy of the Organization of Petroleum Exporting Countries (OPEC), provided oil companies with relatively high margins. Nevertheless, for almost three decades the spread between ROIC and cost of capital for the industry as a whole was negative. Convinced that on a sustained basis the petroleum industry could not deliver a balanced source of income, many companies committed their excess cash to what turned out to be value-destroying acquisitions or other diversification strategies. For example, in the 1970s, Mobil bought retailer Montgomery Ward; Atlantic Richfield purchased Anaconda, a metal and mining company; and Exxon bought a majority stake in Vydec, a company specializing in office automation. All of these cash (or mostly cash) acquisitions resulted in significant losses."
If Icahn started selling shares to someone else, the price would fall. If he sells to Apple itself, much less likely that the price would fall.
Simple example:
company A has 100 shares outstanding and only 1 asset: $100
The value of each share is $1.
It then decides a share buy-back is a good strategy, and buys back 50 shares for $50.
It now has 50 shares outstanding and $50 of assets.
The value of each share is still $1.
The buy-back has no effect on share price.
In any case -- to answer your question -- in your example, it would not have any effect on share price as it seems the company is just a holding company for $100. However, in the case of Apple they have other assets besides the cash (brand, intellectual property) and those assets produce more cash -- so all the cash the other assets generate in the future gets distributed to a smaller number of shareholders.
But you answered as a fact that a share buyback increases share price, and sorry but I don't agree with that point - as demonstrated, for an asset holding company it is not the case.
And for a company with cash generation capabilities there is still a subtle point about whether the company can create more value with the cash than the shareholders; if the company can create more value than shareholders, then removing cash from the company via a share buyback should reduce share price. If it cannot, then removing cash from the company should increase the share price.
Since Apple is simply hoarding the cash, we've already ruled this out -- Apple themselves admits they do not have any way to deploy the cash that would produce more returns than just stashing it in a 0.07% interest bearing account. The question then just comes down to how they disburse the cash (keep vs buyback vs dividends)
In my view you really should qualify your earlier statement that "Buy-backs are more tax efficient because they increase stock price" as it is really not a generally true statement. If it was, I would be in the business of buying companies, then making them use any free cash to buy back some shares from me (my remaining shares then somehow go up in value), then selling the remainder back to the market. Free money, it would be great!
The 80s were well known for corporate mavens buying undervalued companies, splitting them up, and selling the assets individually. The practice resulted in a net profit.
Dividends are taxed in the year of payment, so investors lose some of that gain within a year.
On the flip side, buy-backs can be a poor way to spend cash, as companies might buy back at high stock prices. So the third way to reward the stockholder is to just keep that cash stashed in a bank account -- you get neither taxed (dividends) nor do you buy back potentially inflated shares. However, you have the cash stashed away to do either if you ever need to, or find a good opportunity to (e.g., stock price slump.)