Diversification means buying many types of unrelated stocks to reduce the risk of your portfolio going up in smoke. Some may wonder how this works. I will explore the statistics behind diversification. Let’s say there are N stocks that each have a chance of returning r% with a variance of s% and are statistically independent of each other. If you just buy one stock, your expected return will be r% with a variance of s%. If you purchase two stocks, your expected return will be r_new = 0.5 r_1 + 0.5 r_2 and the variance will be s_new = (0.5)^2 s_1 + (0.5)^2 s_2. In the case that they have the same expected return and risk, your portfolio would have an expected return of r and a halved risk of s/2. Just by buying two stocks instead of one, you reduce your risk by half. Your If you purchased all N stocks, you would still have an expected return of r%, but the variance would be reduced to s/N%. The riskiness is measured by the variance. This is similar to experimental physics, when you do repeated experiments to reduce the error bars.
The key to this working is that the stocks should be totally unrelated. To accomplish this, one should buy stocks in different markets.