I maintain some investments in preferreds, and as others have said, you have to look carefully.
A good majority of preferreds are issued by financial institutions and REITs, although a few others are by other industries/companies, the reason being that they are allowed to place preferred stocks in other areas of their balances sheets that don't impact debt:equity, debt:assets and other ratios. Most have par value of $25/share.
A few points:
Dividends are declared - One item which came up during 2008/2009 en-masse: A company is not obligated to pay a dividend on preferred stock or face a lawsuit, and it's not automatically distributed like a bond interest payment. That's why in press releases, you'll see the company state "and the board of directors declared dividends today of $x/share for the common stock, $y/share for Preferred Class A, and $z/share for Preferred Class G stock". That means that in bad times, they will halt the dividend.
A cumulative preferred stock is one that, if the board decides not to declare a dividend, will keep accruing the unpaid dividend like a taxicab meter. When the dividend is finally resumed, you'll get the unpaid dividends as well that have been accruing. Some preferred stocks have criteria that they can only suspend dividends for up to, say, 5 years before they have to pay it (or go into default on the preferred stock). If a preferred stock is non-cumulative, then if the board suspends the dividends, you're SOL until they resume them. Obviously, they would only do this in times of distress (like 2008/2009), but something to keep in mind.
Usually, I've seen most of the non-cumulative preferreds issued by banks, since they (used to be) are sometimes viewed as "safer" and can get away with this non-cumulative provision, but after 2009, many people are taking a closer look at the non-cumulative provision and might price them lower.
2 case studies on suspended dividends:
FBPRO
A bank in Puerto Rico, with a non-cumulative preferred. Trading at $21/share. It suspended the dividend in 2009, and went through typical bank anguish over the past few years. It's finances are getting stronger, and is returning to profitability. Based on the common stock financials, they look like they could be reinstating the dividends soon (next year or so). When they resume dividends, it'll be paying 8.35% off of a par value of $25/share. At the current price of $21, when the dividend resumes, it'll yield 9.94%, PLUS the market would likely price it to yield similarly with other preferreds at, say, $26+/share, with possible capital gains of $5/share.
Because the dividend is suspended, there are no guarantees when/if the dividend resumes. But like I said, based on their audited public financial statements, their fiscal health is much improved, and even if it takes 2 years to resume the dividend from now, that's a pretty decent return for what I perceive as a 'reasonable risk'. Your analysis and risk profile may vary.
FBS-a
Mid-range, privately-held bank in my hometown, but publicly traded preferreds. It suspended the dividend in 2009. It's owned by a family that runs a large local grocery store chain (i.e. deep pockets). Said family had to pony up a $100 million equity infusion into the bank in 2009/2010 to keep it afloat, so they're obviously committed to seeing it succeed. They expanded into California and were hit with various construction/mortgage loans that went bad. They cleaned up their balance sheet, sold branches in non-core markets (FL, CA) and finances are greatly improved. Their dividend is cumulative. It was trading in the $22s, with a yield at par of over 8% (trading at $22, the yield was over 9%, PLUS it had accumulated dividends of 2 years worth when I bought it). As the bank's finances have improved, the market is valuing the returned dividend as a more likely event, and has since ran up the preferreds to trade above $28.
Note that with a cumulative preferred that suspends dividends, when they resume dividends, there'll be a large one-time dividend of several dollars per share, which will cause the price to drop when it goes ex-dividend.
Call provision - make sure you factor in the call provision on existing shares. If a preferred stock has a callable date of September 2013, and it's trading at $27/share, realize that the company could redeem the stock in September 2013 at $25/share, and you just lost $2/share in capital. Oftentimes the market may price in this possibility with a slightly lower share price, but sometimes it's random and out of the blue. One way to guesstimate this is to look at their total preferreds outstanding. If a company has redeemed older preferreds with higher coupons, and their current preferreds outstanding have various coupon rates, odds are they're going to redeem the highest coupon first (if they are going to redeem it). But, you have to look at when they redeemed them - if it was back in 2003 when times were good, they may not have access to financing now like they did back then if their finances are still struggling. Look at when they redeemed their previous issues, and what the rates were, and what their current preferred rates are. And look at the health of the company.
I have personally experienced the frustration of buying a preferred at a price of, say, $26.50 or $27, only to have it redeemed at $25 6 months or a year later. It doesn't happen too often, but it can happen.
Also, companies can partially redeem preferreds. Your broker might partially call away some of your shares, leaving you with a partial position. So when you look at some preferreds, you might see that the share class was partially called in the past (hint: this is an indication that it could be called again in the near future, so factor that into what price you're willing to pay).
When I look at preferreds I might want to buy, I first look at the common stocks. If the company is somewhat healthy, and has decent financials, I might buy the preferred...but if their common stock is yielding almost as much as (or more than) the preferred, I'll just buy the common. This isn't as frequent these days with the recent equity run-ups, but 2-3 years ago, it wasn't that unusual to find common stocks (mostly REITs/banks) with common yielding as much as (or more!) than the preferreds. Yes, there is more risk of dividend reductions w/ the common - but I factored that into my analysis.