1. Importance of current yield
As the current market is neither super cheap or crazy expensive, there is a good chance of it trading sideways for a long period of time. While buying cheap and good companies is fine, if it offers decent yield, it is an added plus as you get paid to wait.
2. Importance of cash
I'm currently holds more cash than usual because while cash yields almost nothing, if or when there is a crises, on cash allows us to hunt for great bargains.
3. Staying away from leverage
The last thing on my mind is to think about margin call when every thing is falling apart.
4. Consider only companies with very strong balance sheet and pricing power
I will not buy stocks of company that relies heavily on external refinancing as any external shock might force the company to raise dilutive equity. Furthermore, as commodity prices eats into profit margins, it is important to buy companies with significant pricing power who can transfer cost to customer while maintaining its own return on capital.
5. Consider tail risk hedging
I have not implemented this in my current portfolio but is considering doing this. For an over simplistic scenario, see the following:
A December put option with strike price of 108 costs about $2.
The payoff if S&P falls say 25%, (ie from 132 to 99), the payoff is = 108-99 = 9
For a $2 dollar cost, i get at $9 return.*
If i spend about 2% of my portfolio on insurance, should the disaster happen, then the insurance will grow to 9% of my portfolio. The major advantage of this $9 is that it can serve as dry gun powder to hunt for new target during the disaster. The disadvantage is that you'll have a sunk cost of 2% every year cuz you won't have disaster every year.
*Thanks D for your comment. I've amended the payoff accordingly. After the amendment, the payoff doesn't sound as enticing ya.
On a side note, from my understanding of options, usually the only thing that will be mispriced in a option model is the implied volatility. So after many month/years of smooth sailing return, usually implied volatility will be low, thus downside protection can be bought at a cheap price. Conversely, using Black-Scholes model will spit out awkward results for very long dated options when current volatility is high, as it will ignore the compounding effect of retained earnings inherent in the stock market. So Buffett was a put seller during the financial crises.
A sound article that I've read on tail risk hedging can be found here.