Walking Away From A Deal Gets Harder Near The Close

Every live deal file I keep carries a line that reads fees and expenses to date, and it is the only figure in the folder that never falls. By the eighth week of a purchase it has reached five figures and sometimes six, with outside counsel drafting and an engineer still working through the data room. My name is Rosser Newton, buying into privately held energy companies is how I have spent my working life, and that single line has argued with me in every deal I have ever done.

It argues quietly. It never tells me the company is better than I believed a month ago. It tells me how much I have already spent finding out.

Why Does A Deal Get Harder To Stop After Week Six

A deal in week one is a question. A deal in week eight is a schedule, a signing date, a bank that has cleared credit, and four people who have told their families they will be home after the closing dinner.

None of that is evidence about the business. All of it feels like evidence when a finding lands late.

The finding that shows up in week eight is usually the same kind of thing that would have ended the conversation in week one. A customer contract that terminates on a change of control, or a disposal arrangement everybody relies on and nobody papered, or three years of equipment maintenance the seller has been calling discretionary.

Ordinary diligence turns these up on its own schedule rather than on mine.

In week one I would have thanked the owner for his time. In week eight I catch myself asking what it would take to get comfortable, which is a different question wearing the same clothes.

What The Fees To Date Line Actually Measures

That number measures work performed by lawyers and engineers who have already earned it. It measures nothing whatever about what the company will produce next year.

Economists have a clean name for the trap and a long literature behind it. A Nobel went to Richard Thaler in part for describing how badly people handle money they can no longer recover.

The formal treatment in decision theory says the same thing in colder language. A choice gets made on the outcomes still available rather than on what it cost to reach the fork.

Knowing the name of the trap has never once stopped me from walking into it. The protection I use is mechanical instead. When a late finding lands, the memo that goes in front of the investment committee carries the finding and the revised model, and the spend to date figure appears nowhere on the page.

A number that can only argue in one direction does not belong on the page where the decision gets made.

How Do You Keep A Late Decision Honest

An experienced director can argue with me about that, and more than one has. Keeping a real cost off a page shown to the people who authorized the spending looks like managing a committee rather than informing it, and a director who learns later that a page was curated will read the next one with less trust.

My answer is that the figure gets reported in full in the quarterly expense review, where it belongs and where nobody is voting on anything. The decision page carries what the decision turns on and nothing else.

The second practice is older and simpler. Before diligence starts I write down the three findings that would end this deal, and I date the page, which is the same instinct behind a memo written before the money moves. A list made while the deal is still theoretical costs nothing to write and is very hard to argue with eight weeks later, because the man who wrote it had no dinner reservation.

It does not settle every case. A list written in week zero cannot name the thing nobody imagined. It does narrow the argument down to findings that were genuinely unforeseeable, and that set is a good deal smaller than the one people claim.

Where This Rule Breaks Down

Plenty of late findings are price problems rather than deal problems. A receivable that will not collect, a compressor rebuild that was deferred, a title defect on two tracts out of sixty. Those are numbers, and numbers get negotiated.

Treating every one of them as a reason to stop is its own expensive habit. A buyer who walks three times in a year teaches every banker in the state that his signature is theoretical, and the next good company never gets around to calling him.

I have gotten this wrong in the direction nobody warns about. I stopped a purchase in week nine over a customer concentration finding that was plainly a price problem, and I knew it was a price problem while I was saying no.

My own list, written in week zero, said concentration above a line ends the deal. I followed my page past the point where judgment should have taken it back, and the company did fine under somebody else.

So the rule carries a cost and I have paid it. A page written in advance protects me from the weight of the file and exposes me to the thing the page could not foresee, and I have never found a version that does both jobs at once.

What I do now is small. When a late finding lands I put the deal back to the week one question out loud, in front of somebody with no hours in the file, and ask whether we would start this if it arrived this morning.

Half the time the answer comes back yes with a lower price attached, and that is a real answer. The other half I pay the bill, keep the file, and drive home.

I have signed my name, Rosser Newton, under memos recommending each of those outcomes. The only ones I am still sure about are the ones where the page said why.