Even the best investors make mistakes. Some make the same mistakes repeatedly, as a result of behavioral biases. That is the premise of Behavioral Finance, which identifies common, repeated behaviors that impede sound investment decision making. While we can strive to identify which behavioral missteps each of us is prone to, it often is too much to ask of us humans to overcome those tendencies. Instead, the best approach in overcoming these flaws is to use disciplined valuation approaches, such as stock prices as multiples of book value, revenues, or trailing earnings.
Being human myself, I keep making mistakes despite more than 25 years as a professional investor. I tend toward "confirmatory bias," which is to seek information that confirms my existing view instead of seeking opposing views that might expose flaws in my reasoning. The one that continues to bedevil me is my reluctance to admit a mistake and sell a losing position. A Wall Street adage is to "cut your losses (promptly) and let your profits run." Too often, I do the reverse, citing a different adage (by a former colleague), "You never go broke taking a profit."
There are two types of losses on a stock. When a stock declines from the purchase price, it could be the result of negative news from the company that might be at odds with the original reason for buying the stock. In this case, then it probably makes sense to take the loss. The other type is for no company-specific reason, other than perhaps an overall market decline. In this case, if the stock was at an attractive price in the first place, I'd be inclined to add at a lower price--or at least hold.
When speculating on an instrument or a commodity (or even a stock), however, there isn't the valuation support to justify holding a losing position. In that case, the challenge is to get a sense of whether it is a temporary decline or the beginning of something longer and more substantial.
A number of investors use "stop-loss" orders to sell out of losing positions at preset prices, in order to limit losses to a specified percentage. (Abrupt price declines might skip over the stop-loss price and result in an even larger loss, however.)
I have had a longstanding bias against using stop-loss orders. , If I like the price I paid and the stock price drops, I'm inclined to buy more as long as the valuation is better and the fundamental case is the same. But after a sizable recent loss on a speculative instrument, I have reconsidered using stop loss orders and now recommend them in that type of situation. I'm not ready to advocate their use in long-only investments in stocks that are based on solid financial conditions, good business prospects, and favorable valuation.
Successful investing requires continual efforts at improvements in market knowledge, as well as in trading or investing techniques. I'm still learning.