I'm going to assume that you want to be invested all the time and each trade consists in selling a security and buying another one (similar to your example).
How much commissions you are willing to pay depends on several factors, but one way to think about it is as follows. You have a position in stock A and you want to switch to stock B because you think it will perform better. If you think there's a good chance (>50%) that B will outperform A by more than x% then you can happily pay up to x/2% commissions and still make money over a long time horizon.
If you like formulae, one way to express it is:
expected_return = hit_rate * (win_loss_ratio - 100%) - commissions
hit_rate is the number of winning trades vs. losing trades. A hit rate of 55% means that out of 100 trades, you make money on 55 and lose on 45.
win_loss_ratio is the average gain vs. average loss. If on the 55 winning trades you made $110 on average and you lost $100 on average on the 45 losing trades, you ratio is 110%.
Example: if you tend to be right 51% of the time (hit rate), and gain 110% more than you lose on average (win loss ratio), you can see that your expected profit is:
5.1% - commissions, so you could pay 2.5% commissions on entering and closing the position and still make money*.
Unfortunately, common sense, statistics and numerous studies tell us a sad truth: on average, people have a hit rate of 50% and a win/loss ratio of 100%. Which means that their expected profit per trade is
0% - commission.
Based on that crude observation - unless you can prove to yourself that you are better than average - you should aim at reducing commissions paid to your broker as much as possible through:
- low commissions arrangements: in your example you pay 0.5%, which is quite high. You should aim at 0.2% or less (if you pay an amount per share, which is common in the US, you should avoid penny stocks)
- low amount of trading (turnover)
* 51% and 110% are not random numbers, they correspond to the results of the top 15% (professional) managers in a research paper using a sample of 215 funds managing $150bn.