NEW YORK (TheStreet) -- There are more than 900 publicly traded banks and thrifts in the U.S. including Astoria Financial (AF) . Which are solid performers? It's a complicated question.

The place to start is return on tangible common equity, or ROTCE. The better the bank, the higher the ROTCE. Why? Because banks pay out a percentage of their earnings as dividends and retain the rest to support asset growth. Higher ROTCE banks can grow faster or pay higher dividends or do both than low ROTCE banks.

Of course, banks that use aggressive leverage or take on excessive risk to boost ROTCE can end up destroying shareholder value, a painful lesson that bank stock investors learned in recent years.

But the story doesn't end there. Every bank stock has a price, and price reflects operating performance in theory but not necessarily in practice. So price can trump performance; at the extremes, a great bank may be just too expensive to own, or a mediocre one may be too cheap not to.

Which brings me to Lake Success, N.Y.-based thrift Astoria Financial. In a Jan. 2 TheStreet article, Meaningful Upside' for Astoria Financial in 2014, Says KBW, analyst Brian Kleinhanzl raised his 2015 earnings estimate for Astoria to $1 a share from 85 cents, based on a repositioning of Astoria's balance sheet that should enable the company to benefit from increases in interest rates. His 2014 estimate remained unchanged, at 70 cents. He also raised his 12-month price target to $17 from $13.

Should an increase in projected EPS lead to an increase in expected share price? Sure, unless the stock you're looking at was overvalued to begin with. It's hard not to think Astoria is overvalued given its longstanding mediocre operating performance. The bank's market cap at the market close on Monday was $1.36 billion.

For the 84 banks and thrifts with market caps between $500 million and $2 billion, the average price/tangible book value multiple is 1.9. Astoria's is only 1.2. So Astoria must be cheap, right? Not so fast. The average ROTCE year to date through the third quarter for this peer group is 11.9%. Astoria's is only 4.9%, and it hasn't been materially above this level since the beginning of 2011. So Astoria isn't cheap. It has a relatively low trading multiple for very good reasons.

Astoria's assets at the end of the third quarter were $16 billion, 32% below the fourth-quarter 2004 peak of $23.4 billion. Its assets in the first quarter of 2000 and the fourth quarter of 2008 were essentially the same, astonishing when you consider the robust asset growth many banks and thrifts delivered during this period. Sure, that robust growth turned out to be problematic or fatal for some institutions, but isn't 32% asset shrinkage awful?

While Astoria's asset shrinkage has slowed recently, it still has grown assets only three times over the past 12 quarters, and only once at an attractive level. The thrift has been shrinking its securities portfolio, shrinking its one- to four-family residential loan portfolio (not an especially lucrative loan category) and growing multifamily, commercial real estate and other loans, all in an attempt to boost its net interest margin. NIM has improved relative to the lows of 2007, but at 2.28% during the third quarter it's still slightly worse than the 2.40% way back in the first quarter of 2000.

Both are horrible in absolute terms (in fairness, this is a problem for many thrifts, not just AF). But here's the bigger problem: Astoria's non-interest expense has grown even as assets have shrunk. Its overhead is much too big for its current asset base. That will be a huge, ongoing drag on its returns on assets and equity.

TheStreet Recommends

If Astoria hits Kleinhanzl's $1 EPS estimate in 2015, that would give the bank roughly an 8% ROTCE in 2015, more if the thrift increases its dividend. That's better, but still far from good. Note that Kleinhanzl revised his 2015 estimate upward by 15 cents and his price target by $4. How does one map into the other? I have no idea.

If you assume that the entire 15-cent increase is paid as dividends, 4% dividend growth and a 7% discount rate gets you $3.51. But those are three very aggressive assumptions for Astoria. Kleinhanzl predicts 43% EPS growth for 2015. His 2015 estimate is far from a sure thing, but even if Astoria hits it, where does EPS growth come from after that? More net interest margin expansion? Asset growth?

I have to believe the 48% increase in Astoria's share price in 2013 is the real reason for Kleinhanzl's optimism. But Astoria's shares traded at 83% of tangible book value at the end of 2012. Enough investors considered it irrational for a thrift to trade below its theoretical "liquidation value," even one with a terrible ROTCE.

Astoria's operating performance in 2013 hasn't been that much better than that of 2012, and thrifts in general, several with operating performance far superior to Astoria's, appreciated significantly, if not as much as Astoria. Even with its massive 2013 price appreciation, the $1.35 billion Astoria spent on share buybacks from 2001 to 2006 appears to have been wasted; the weighted average repurchase price over this period was $23. Astoria hasn't repurchased shares since 2008, even when they traded below tangible book value, despite the availability of excess capital.

One final point: Even with AF's depressed late 2012 valuation, short interest in its shares in mid-December 2012 was 11.6%. Short interest fell steadily during 2013, and now stands at a much lower, although still high at 4.9%. So there's some skepticism in the market about where Astoria's price might go from here.

I've just presented what I consider several useful facts about Astoria Financial. Maybe Kleinhanzl believes other facts matter more, and maybe those other facts will propel the stock to $17 per share and beyond in the near term. Or maybe facts don't matter. If a plane runs out of fuel, what happens to it is independent of its passengers' hopes.

But in the stock market, investor hope can be enough to sustain prices, at least for awhile. An investor may choose to bet on hope, but mistaking that for a bet on fundamentals can lead to poor returns in tech stocks and bank stocks alike.

This may be the year that bank and thrift investors relearn this lesson.

At the time of publication, the author had no position the stock mentioned.

This article represents the opinion of a contributor and not necessarily that of TheStreet or its editorial staff.

Harvard Winters is a former financial services sector investment banker who now writes research on banks and thrifts.