Here's a quick&dirty way to go about insurance:
Take the probability of event occurring over given timeframe (say, 10 years). With Identity theft, this includes:
- You compromise your password by writing it down to a stickynote, that gets stolen
- Your computer gets rooted by software / hardware means AND the account information gets stolen
- One (any) of your service provider with your CC (or close proxies to that) gets compromised
- Your bank getting compromised
Let's say, for the average Jane the sum of these probabilities P(compromised) to be around 5% within a 10 year timeframe (YMMV, obviously). Let's further assume, that half of the compromised identities gets exploited.
P(your bank account's get exploited) = P(compromised) * P(exploited) = 2.5%
Let's further assume, that exploiting average Jane, taking her savings account, is worth 10 grands.
How much is Jane expected to loose on average over a 10-year period?
L = P(your bank account's get exploited) * Take = 10000 * 0.025 = £250
Now, according to first Google hit on identity theft insurance,
Identity theft protection generally costs between £3.75 and £6.99 a month or between £45 and £84 a year.
Hence, total cost of solution, for a 10-year period, for the most minimal insurance, which I'm not even sure whether it covers backing actual money:
3.75 * 12 * 10 = £450
In this case, Jane is probably better off without it.
Using the data above, at what apriori probability is this actually worth it?
P(your bank account's get exploited) * Take > 450 => Take = 10000
P(your bank account's get exploited) > 4.5%
In conclusion, you might want to look into Identity theft insurance, if either your estimated probability of having your account compromised, or the amount of potential take outweighs the cost of ownership.
Hope this makes sense, let me know if you have any questions.