British bank Barclays has been fined £26 million ($43.7 million) by financial regulators for attempting to rig the price of gold for its own profit. It is the latest scandal to shine a light on the sketchy practice of self-certification, which only encourages actors to intentionally misreport the data out of self-interest.

The incident in question occurred in June 2012, the day after the bank was fined a record £290m ($488.6 million) for helping rig LIBOR, a global interest rate benchmark based on self-reported financial industry lending costs. LIBOR is used by mortgage and credit card providers to determine interest rates, and is tied to at least $350 trillion of assets in derivatives and other financial products.

BBC explains how the gold scheme worked: "If the price of gold was above $1,558.96 [per ounce at the end of the day], Barclays would be required to make a payment of $3.9 million to its customer. But if the price was below that benchmark, Barclays would not have to make the payment."

Daniel James Plunkett, a director on Barclays' precious metals trading desk, created fake orders and submitted them to the gold fix — a twice-daily publication of the gold price based on market activity and banks' current orders — "with the intent of pushing the price of gold below $1,588.96, which he succeeded in doing."

The result? Barclays dodged the $3.9 million tab, and Plunkett scored a profit of $1.75 million for the bank.

Lucky for the customer, Barclays was caught. But we have no idea how many times banks have done this and not been caught, because the data is self-certified and not subject to significant oversight.

Barclays' fine was considerably larger than the $1.75 million it attempted to gain, but that's beside the point. After all, Barclays had been fined almost $500 million literally the day before for its part in the LIBOR rigging scandal. The truth is that the size of the fine doesn't matter in preventing these kinds of behaviors — it's basically the cost of doing business.

The real problem is the practice of self-certification, which only encourages banks to lie and cheat. An honor system will no longer suffice. What we need instead is independent oversight to guarantee that banks are not lying when they submit data to determine the gold fix, oil benchmarks, and interest rate benchmarks like LIBOR.

Self-certified data is notoriously problematic, and not just in finance. Let's take a topic close to a very large number of people's hearts, so to speak: penis length. Studies in which a medical professional measured men's anatomy suggest the average length is in a range between 5.08 and 5.71 inches. Studies on penis length based on self-reported length suggest the average length is in a range between 6.14 and 6.54 inches.

This is human nature. When we have an incentive to exaggerate or lie with little chance of being found out — either to bolster our egos or to cream an easy profit — we have a propensity to do so.

If we were to give regulators and auditors the power to open firms' books to investigate all such self-certified data submissions and any conflicts of interest, and mandate prison time as the minimum penalty for making fraudulent submissions, then we would quickly see this problem go away.

Editor's note: This article has been revised since it was first published in order to more clearly include proper attribution to source material.