This is a misunderstanding of the rules of probability. For systems like this, past results have no bearing on future results. The expected profits for April are not related to what happens in March. (With the caveat here that since this is a business things are a bit different. If we actually get a loss for the month, then that would be carried forward into April, but if it just breaks even then April is unaffected.)
To get a better grasp of this, lets assume that when you have a child you have an equal probability of having a boy or a girl. I already have two girls, what is the probability that my next child will be a boy?
Answer: The probability of having a boy is 50%, the same as it was for the first child.
Of course statistically this is known as the Gambler's Fallacy. Yes we could stay above expected profit forever since each win/loss is independent. We could also stay below the expected profit for an infinite amount of time.
As a practical matter expecting a large amount of independent die roles to ultimately be closer to the expected probabilistic outcome rather than further is how all casinos generate money and long term why you can't beat the house. In the long run profit will return to the expected value based on expected probabilities.
You are still wrong. One does not expect the future outcome to trend back toward the expected value set several months ago, one expects the future outcome to follow the trend line starting at today.
As an example, lets say we are going to flip a coin 20 times. At this point we predict 10 heads, 10 tails. We then flip the coin ten times, and despite probability we end up with 10 heads. What you are saying is that you expect the next 10 coins to come up tails, putting us at the original expectation. What I am saying is that for the next 10 coins I expect 5 heads and 5 tails, so my new expectation is we will end up at 15 heads and 5 tails overall.