Fool analyst Rich Smith recently discussed business development company American Capital's (Nasdaq: ACAS) decision to repurchase millions of its own shares at the end of  last year -- for those of you wondering, American Capital is also the parent company of the popular mortgage real-estate investment trust American Capital Agency (NYSE: AGNC). Rich found the decision imprudent, given that the company can't afford to pay a dividend and is still recovering from the massive $4-billion-plus losses it suffered in the financial crisis.

As often happens with our articles, a handful of readers disagreed with Rich's assessment. They argued that the decision was good for shareholders because American Capital's shares are trading for a fraction of book value, a commonly used valuation metric for business-development companies. According to one reader, "[American Capital] has been trading around 60% of book value. Buying back shares is giving way more value to shareholders per dollar than paying a dividend would. Would you rather have [a $0.60 dividend], or buy a dollar for [$0.60 with a share buyback]?"

Given my affinity for financial stocks -- not to mention a recent interest in business development companies -- I've decided to jump into the fold. And it so happens that I, like Rich, believe American Capital's decision represents an imprudent allocation of capital. Before getting to the reasons why, however, I wanted to discuss buybacks more generally for those of you who are new to the debate.

A brief background on share buybacks
It's hard to dislike the idea of a share buyback. By repurchasing their own stock, companies reduce the number of pieces that the corporate pie is divided into, leaving existing shareholders with a larger slice of everything from net assets to earnings. Similar to a typical stock purchase, however, the prudence of a specific repurchase program comes down to timing: A company wants to buy low and sell high.

Berkshire Hathaway's (NYSE: BRK-B) decision to repurchase shares at the end of last year provides a great example of the thought process behind a profitable buyback. In announcing its decision, Berkshire's board of directors stated their opinion that "the underlying businesses of Berkshire are worth considerably more than" the current price. Its shares at the time were trading at 1.02 times book value, or about 50% below their historic average. The shares have since increased in price by nearly 11%, providing the company and its existing shareholders with an annualized gain of 34%.

Unfortunately, Berkshire's experience in this regard appears to be an exception to the rule. My colleague Morgan Housel has written about this on numerous occasions. He disclosed the hundreds of millions of dollars that Circuit City, Countrywide Financial, and Citigroup (NYSE: C) sunk into ill-timed buyback programs in 2006. He discussed how banks like Bank of America (NYSE: BAC) and Goldman Sachs spent billions of dollars buying their own shares at the height of the housing bubble, only to watch as the market collapsed soon thereafter. And in an article last year, Morgan offered a revealing chart that shows the inverse relationship between share buybacks and subsequent returns.

Where does American Capital fall?
At first glance, it's tempting to conclude that American Capital's recent repurchase program has more in common with Berkshire's than it does with Circuit City's or Bank of America's. In the first case, the company's share price appears closer to the market bottom as opposed to its top, as its shares have fallen by more than 80% since mid-2007. And in the second case, its shares were trading for 36% less than book value during the program -- that is, American Capital's book value per share is $11.92 whereas its average repurchase price was $7.61 per share, according to Yahoo! Finance.

However, a closer look suggests this conclusion may be premature. Remember what I said about the strategy underlying a share buyback? You want to buy low and sell high. American Capital appears to, well, have done the opposite. According to a recent quarterly filing (page 67): "In April 2010, we completed a registered direct offering of 58,300,000 shares of our common stock to a group of institutional investors at a price of $5.06 per share" (emphasis mine). In other words, the company effectively sold 58.3 million shares for $5.06 a share in April of 2010 only to repurchase 17.5 million of them a year and a half later for an average of $7.61 a share. So no matter which way you slice it, it's hard to disagree with Rich's assessment of American Capital's buyback decision.

Foolish bottom line
At the end of the day, I have to admit that I strongly dislike share-buyback programs -- unless someone like Warren Buffett is calling the shots, that is. While their stated purpose is to increase shareholder value, I believe their true purpose is more typically to mask executive stock grants by counteracting what would otherwise be a dilutive form of compensation. If you disagree with this or my analysis of American Capital, I encourage you to say so in the comment box below.

Otherwise, I suggest that you take a look at a recent free report by our analysts here at the Fool about stocks only the smartest investors are buying. It profiles a too big to fail bank that is roaring back to relevance and making shareholders rich along the way. And it also discloses the identity of a smaller financial institution that Warren Buffett would have likely been attracted to in his younger years. To access this report while it's still available and before the rest of the market catches on, click here now -- it's free.