The question is impossible to answer, because an important information is missing.
The missing information is: What is the probability that a random coin is biased? More precisely, what is the distribution of biased coins with different biases in the initial population of the coins? What is the prior probability that the coin was biased, before you run the experiment?
To see why this information is relevant, imagine that the experiment was repeated in two different countries. In country A, biased coins simply do not exist. In country B, only 1/3 of coins is fair, 1/3 is head-biased, and 1/3 is tail-biased. In both countries the experimenters run the experiment, in both countries they receive the same 40:60 outcome. But because of the difference between the countries, the two experimenters must interpret their results differently.
In country A, where biased coins do not exist, we must reason according to Sherlock Holmes that "when you have eliminated the impossible, whatever remains, however improbable, must be the truth". Biased coins are impossible, therefore the truth must be that a very improbable event happened with a fair coin.
In country B the initial odds of fair coin versus tail-biased coin are 1:1, so after the experiment, the resulting odds are something like 1:50. (I am not sure about the exact number. Also it depends on how strongly tail-biased those tail-biased coins are. If they are so biased that they provide tails in 90% of flips, then receiving a 40:60 outcome with such coin is even less probable that receiving a 40:60 outcome with a fair coin.)
In what country did you do your experiment? I guess it's somewhere in between the example A and B countries: there are biased coins, but they don't make 1/3 of the coin-population. But you have to make an estimate. Do you think that 1 coin in 10 is biased? Or 1 in 1000?
Calculate a probability that you have achieved given result with a fair coin. Probability that a random coin in coin-population is fair × probability that flipping a fair coin 100 times will give you 40:60 outcome.
Calculate a probability that you have achieved given result with a tail-biased coin. Probability that a random coin in coin-population is tail-biased × probability that flipping a tail-biased coin 100 times will give you 40:60 outcome. (You may do this calculation multiple times, for different degrees of bias.)
The ratio of these results is the ratio of posterior probabilities of having started with a fair coin vs having started with a tail-biased coin.
(For a more complete calculation you should also include probabilities that you made a mistake in calculation, or that the person who flipped the coins has a skill to make a fair coin land on given side with higher probability. Sometimes these probability are not insignificant compared with probabilities of achieving the result by regular methods.)
Without this data you can't calculate the real probability, because you don't have the relevant inputs. You can only calculate "something", which may be called a traditional interpretation, but is incorrect. But it's traditional, which means that many people make this mistake. (Achieving an incorrect but credible result may be preferable to a correct calculation containing too many unknown values.) The correct method of calculating probabilities is called "Bayesian".