I short quite a lot at my current shop. I'd split my shorting into 4 buckets:
Low quality companies that are misunderstood:
These are typically misunderstood because some risk factor in the business model isn't fully appreciated by the market, or because analysts get too optimistic and project out constant (or accelerating!) growth and/or expanding margins. Example of this is retailers that are dependent on store growth who end up cannibalising sales which isn't in consensus projections. Sometimes primary research can almost be too easy (e.g. Restoration Hardware, which was priced as a premium retailer but ran endless promotions in 4Q15). But it's usually pretty hard. Ideas typically come from knowing the companies in your sector very well and understanding when expectations get out of hand.
Typically more macro-based, e.g. trying to get short exposure to commodities cycle by shorting commodities-related companies; exposure to US minimum wage inflation by shorting labour-intensive businesses; etc. For me, this tends to be a by-product of macro research I do for my long book.
Frauds and promotes:
Typically smaller-cap and a lot of work, but some people are very good at this (Muddy Waters, Citron). Idea generation from various screens that indicate aggressive accounting (e.g. working capital, tiny auditors for relatively large companies, changes in accounting policy, etc.), and following lists of known stock promoters and crooked managers (but that's typically at the small and micro cap level). These are rarer for me, because to earn a good return on effort, you need fairly concentrated position sizes, which I don't like (see Position Sizing below).
Hedging / pair trades:
Sometimes we short to hedge specific risk factors in the rest of the portfolio, e.g. currency, macro, volatility, or just to manage our net exposures to a specific target. These tend to generate, but they enable us to put on more long exposure if we have long names that we have a lot of conviction in. Obviously you'd rather manage your net exposure by having "alpha-shorts" but sometimes you just don't have enough ideas.
I'll expand on this below, but timing and trajectory is of the utmost importance to me when shorting. I tend to maintain a long list of names I'd like to short, and wait for opportune moments to build the positions. It means letting a lot of things go pass, but that's part of prudent investing in my view.
I follow a few rules when shorting:
Never short high quality companies:
Mainly because the sky is the limit on valuation when there is flight to safety, yield compression, or investors chasing quality growth. This is stuff like consumer staples and IG bonds in the last 2 years, and growth-y names like Under Armour (yes, some people made money shorting but plenty more got burned).
Never short "story stocks":
Similar to (1), when people get enamoured with a blue sky story, they can justify any valuation, even when fundamentals are totally out of whack vs. price. See: 3D printing companies. Timing the peaks on these doesn't fit my style but I'm sure there are people out there who do it well. Generally, going against the crowd isn't easy, and being 'contrarian' doesn't necessarily mean being smart (probably the opposite in most situations).
The route to your target price is absolutely crucial. If a stock goes from $50 to $5 via $100, you'd have been stopped out way before you got to your target price and lost money. Shorts that go up work against you doubly, because gross exposure increases which means you often have to lock in losses as you trim your exposure to manage risk. I'll come across stocks that I think are long-term shorts, but pass because I think I'll get stopped out at a loss. This is why I find timing your entry on shorts is more important vs. longs (where you can double down if you're early). Options can be a way around this, but I find writing calls unattractive and expected volatility will usually work against you if you're buying puts in these situations. You're better off riding a short that's already on a downward trajectory (i.e. some aspect of the short thesis is fundamentally proven), particularly if it's a zero (or close to it), since the return from $100 to $10 vs.$70 to $10 isn't all that different.
Shorting on valuation alone without a catalyst is a death wish because of (1) and (2). Also, I typically want my shorts to work a lot faster than my longs. This relates to (3). Since most stocks have positive beta and the market tends to rise over long periods of time, you are more likely to get burned if you hold on to your position for a long time. Also, the absolute return on a short tends to be lower (because the maximum is 100%), so you need to consider IRRs as well.
I suppose this is self-explanatory, but short squeezes are painful. Again, because you will get stopped out even if you're right (you can tell avoiding this is a common theme with me). Thankfully I haven't experienced one first hand (yet). I tend to pay more attention to the borrow cost rather than % float that's shorted, because I think borrow is a much better indicator of true liquidity. See: Volkswagen / Porsche. Fraud-y microcaps can be very easy to squeeze too. See: KBIO.
Buyouts are probably the worst outcome for a short, because you get forcefully closed out, and will never make that money back, even if you're right and the acquirer ends up writing down the entire acquisition years later. Moral victories are nice but you can't charge 2 and 20 for them. Many low multiple value traps are good shorts, but I always make sure I don't have too many in my short book because they can look cheap to some buyers. See: Green Mountain Coffee Roasters, Diamond Resorts.
You typically want your individual shorts to be smaller than your longs, because their exposure naturally increases when you're wrong. This is one of the more irritating aspects to shorting, because you often expend a lot of effort for a not-very-large position size, so on an individual basis, shorting has poor return on effort. But from a portfolio perspective, it lets you be more long, so it increases your return on effort for your longs.
Right, this ended up being a far longer post than I intended, so apologies if it's a little rambling or disorganised. It's most definitely not a comprehensive guide to shorting, but I thought I'd just share some of the aspects from my process.
Mod Note (Andy): This very helpful comment by thewaterpiper was originally posted 7/25/16 in the thread What is the Shorting "Research" Process and received 20 silver bananas, so we thought it deserved its own spot on the frontpage for those who may have missed it.