Whether age predicts behaviour
The statement: they have been around for two years, so they are fine. Age is evidence. It is much weaker evidence than the confidence people put in it.
Bucket: DEPENDS
It depends on what the age is measuring, whether the account has been continuously the same hands, and whether the value at stake has changed. Age is a measure of duration, not of intent.
What age genuinely tells you
That the account survived a period of time without being removed. That is real. It means transactions happened and complaints did not accumulate past whatever threshold applies. For an account with a long history and steady activity, that is a signal worth having.
It is a signal about the past behaviour of whoever was operating it. That is the whole of it.
What it does not tell you
- Who is operating it now.
- Whether the account was bought.
- Whether the operator's situation has changed.
- Whether the incentives have changed.
- Whether age was manufactured by low value activity.
The failure mode age makes worse
The pattern where a long standing account trades on its record for a short period and then stops honouring anything. The accumulated reputation is the asset, and the value of spending it rises the longer it has been built. Age does not protect against this. Age is the precondition for it.
Which flips the usual reading. The oldest, best regarded account is the one with the most to spend, and there is no way to distinguish an account that is about to do this from one that is not.
What people get wrong
They read age as character. What accumulated was a record of transactions, not a demonstration of ethics. It is compatible with a person who behaved well because it was profitable to.
They also fail to notice discontinuity. An account that went quiet for months and came back is not the same evidence as one that ran continuously. The gap is where a handover would sit, and the age number does not show gaps.
What to look at instead
- Recent activity rather than total duration.
- Whether the tone and detail of communication is consistent with older records.
- Whether behaviour changed after a quiet period.
- Whether the amount at stake is large enough to be worth spending a reputation on.
None of these are proof. They are better questions than how long, and they cost the same to ask.
The other direction
A new account is not evidence of anything either, and treating it as a warning is the same mistake with the sign flipped. Every long standing account was new once, and somebody transacted with it then. New means no record, which is an absence of information rather than bad information.
The reasonable response to no record is to size the exposure to the uncertainty, which is a decision about amounts rather than about people. That is available to you and does not require reading anybody character.
What would move this into ALWAYS or NEVER
Nothing available. It would need a way to bind an account to a continuous operator, which is exactly what an anonymous system does not provide. So this stays conditional, and the conditions are the ones listed above.
Questions people actually ask
Is an old account safer than a new one?
It has a record where a new one has none. That is an absence of information rather than proof of anything, and a long record is also the asset somebody would spend.
Can an account be bought?
Yes, and nothing visible from outside distinguishes a sold account from one still in the original hands.