JP Morgan Did Not Measure Twice and was Defrauded by 31-year old

Charlie Javice swindled $175 million from world's biggest bank.

Neglecting to Measure Twice. Why are so many smart people not smart enough to realize what they don’t know? Probably because they want to be admired as the “smartest” or the “fastest” to make a judgement. Many times they are correct, but situations in which there is no second measurement are problematic since these constructs entail Iceberg Risk. As we know, 90% of an iceberg is underwater and not visible. Learn to identify situations that look great on the surface but do not hold up to that second measurement since this is where your dreams will likely hit that iceberg. Here is the story of how JP Morgan rushed into an acquisition, did not measure twice and was burned badly.

Neglecting to Measure Twice is one of the Seven Deadly Stupidities.

By most measures, JP Morgan is the largest financial institution in the world. It has more than 250,000 employees and does business in almost every country on the planet. JPM has successfully acquired and assimilated multi-billion acquisitions like Chase Bank and Bear, Stearns. But it is the relatively small acquisition of Frank Financial Aid that is instructive to examine.

Frank was built upon the fact that 85%+ of college-bound students apply for financial aid using the Free Application for Federal Student Aid (FAFSA). I’m sure you have heard of it. The problem Frank attacked was the mind-numbing complexity of filling out the FAFSA correctly. Many say they would rather fill out tax returns than work on the FAFSA. 

The FAFSA is the cornerstone document used by federal and state agencies that award grants and loans, as well as the document used by two-year, four-year, vocational, and any other college to assign aid. The average total awards driven by the FAFSA are more than $20,000 per student. Real money.

Frank’s basic service was free, and it helped users fill out the FAFSA, similar to popular tax preparation software like TurboTax. It offered various levels of paid consulting services to help applicants maximize the value of awards. 

For JPM, the rationale for the acquisition of Frank was that it provided JPM greater reach into the population of college-bound young adults. By acquiring this client base, JPM would have the opportunity to penetrate this Generation Z population and have them start using JPM’s credit cards, bank accounts, stock-trading functions, auto loans, home mortgages, and more. Seemed to make a lot of sense at the strategic level for JPM.

Prior to any acquisition being finalized, there is a period of due diligence. Rather than simply believing whatever the seller says, the due diligence process provides the buyer with the opportunity to review detailed information that supports the seller’s claims. For example, if the seller claimed $100,000 in sales last month, the buyer would review all the invoices for the month and see if they add up to $100,000.  Further, the buyer would randomly call some clients to verify goods were sold and delivered and the invoices were valid.

Due diligence is essentially an investigation by the buyer (JPM) of the seller (Frank). Having been part of many due diligence processes myself as a buyer and as a seller, I can attest to the simple dynamic that exists in every diligence project:  The seller wants to share as little as possible and the buyer wants to know everything possible. This dynamic exists for the simple reason that once serious diligence starts, the buyer and seller have already agreed on price and other key terms (e.g., who gets employment contracts).

So, for the seller (like Frank), diligence is all downside.  Think about when houses are bought and sold. A price is agreed on and then the buyer sends in the inspectors and comes back to the seller and says, “When we made our offer, we didn’t know that the heat does not work in the upstairs bathroom. It will cost us $25,000 to fix it, so let’s reduce the price by $25,000.” In this case, the house buyer did her diligence and measured twice before cutting once.

In the case of Frank, JPM thought it was acquiring a profitable business with 4.2 million new accounts as targets for the menu of JPM services. These accounts were listed in a database and included name, address, and other personal information on each account. After the acquisition closed and $175 million changed hands, JPM discovered a disturbing fact about those 4.2 million accounts: more than 90% of them were fabricated. Oops.

Soon after the merger closed, the bank took its shot and sprayed a portion of Frank’s customer list with solicitations. Of 400,000 outbound emails, only 28 percent arrived successfully in an inbox, compared with the usual 99 percent delivery rate. Moreover, just 103 recipients clicked a link to Frank’s website.

It was, as the bank put it in its legal filing, “disastrous.” New York Times

As the JPM-Frank drama unfolded, we learned that the Frank CEO and Forbes 30-under-30-star Charlie Javice went to extraordinary measures to perpetrate the fraud at Frank. Javice hired a data sciences professor to create the fake accounts. She paid the professor $18,000 to generate the data. 

The question is: Prior to the deal closing, did JPM measure anything? Frank was only in existence for a few years at the time of the JPM deal. There are about 1.6 million college applicants per year, so if Frank had 4.2 million accounts, it would have meant that almost every college applicant since Frank was founded was using Frank’s service. That should have been a threshold question at the onset of any diligence activities and asked before major deal terms were constructed.

Just as troubling is the lack of rigor in the JPM diligence process. When I have been involved in diligence of large databases like the 4.2 million accounts at Frank, the process was straightforward:  hire a diligence team from a national consulting firm, have the statistician create a representative sample of the 4.2 million accounts, and hand the sample over to a call center or mailing house to verify the existence of the sample accounts. In this case, a 1% randomized sample would have been 42,000 accounts and maybe a couple of weeks of work for a professional call-center to verify each one. If, for example, only 50% of the sample could be verified, then the acquisition process would come to a full stop to determine what the heck was going on and the question would be, “Are we really buying 4.2 million accounts?”

Fraud is always a difficult one to pin down. Fraud must be “willful and intentional.” Hiring a professor to generate fake accounts, well, that’s fraud.

Many times, there is a delicate dance between the buyer and seller during negotiations and diligence in which the buyer does not want to come on too strong and scare away or drive the seller to a different buyer. This is common behavior, especially in a competitive bidding process. It’s stupid.

If your questions as a buyer are going to anger or upset the seller, then so be it. If the seller walks because of your approach, believe me, the deal would have fallen apart for some other reason soon enough.


or to participate.