On the Origins of Stormy Daniels’s $130K Settlement Payment, and the Probability of the Disbursement Records Matching the Settlement Amount by Random Chance

This blog post is a continuation of a Tweet-thread I posted earlier this week about Trump’s settlement/nondisclosure agreement with Stormy Daniels, and the questions surrounding the original source of the $130K that was used to make the payoff.

Originally, I had been planning to continue with additional updates on the issue in tweet form… until at some point I realized it would’ve gone on to like tweet 736/736 to actually do it. So in order to spare everyone that, I decided it was time to dust off my poor neglected blog instead, and do a real post for once.

Brief background on the Stormy/Trump Nondisclosure Agreement & Settlement

• Stormy Daniels (real name Stephanie Clifford) and Donald Trump (fake name David Dennison) had an affair in 2006-2007, after meeting at a Tahoe golf tournament.
• The affair was partially reported on in 2011, but never confirmed.
• During the 2016 election, beginning sometime in late summer of 2016, Stormy’s attorney Keith Davidson approached the Trump camp (presumably meaning Michael Cohen, Executive Vice President of the Trump Organization and Trump’s personal attorney) to inform them of Stormy’s plan to speak out about the affair.
• Cohen countered with an offer of a confidential settlement and nondisclosure agreement instead.
• Stormy accepted, and an NDA was negotiated between the parties; the total pay out to Stormy would be $130K, and in exchange she would agree not to disclose any details of her affair with Trump.
• In October 2016, less than a month before the election, the deal began to fall apart when Trump failed to make the agreed-upon settlement payment.
• Consequently, Stormy Daniels spoke to reporters about finally going public with her story, apparently concluding that if Trump was never going to actually pay her, so she might as well get paid for her story by telling it publicly. According to a Slate article published in January 2018, this occurred in about mid-October:

Daniels said she was talking to me and sharing these details because Trump was stalling on finalizing the confidentiality agreement and paying her. Given her experience with Trump, she suspected he would stall her until after the election, and then refuse to sign or pay up. …

I told Daniels that Slate did not pay sources but encouraged her to come forward without compensation. I proposed interviewing her on Trumpcast and writing her story. She never said yes and never said no. Late in the discussion, I asked a Slate colleague to help me verify her account. We both spoke to Daniels and to Gina Rodriguez, a former porn actress turned agent, who Daniels was using to negotiate with media organizations. I gathered that Daniels was also discussing going public on Good Morning America. At one point she considered holding a press conference in Dallas, where she lives.

• On October 18, 2016, one such story was published by The Smoking Gun
• But subsequently, “about a week before the election,” Stormy went silent, having presumably taken the payout from Trump and agreeing to the NDA

The Wall Street Journal’s Coverage of the Settlement Payment

The Stormy story disappeared into nothing in the final weeks of the election, and essentially disappeared from view for the next 14 months. Then, in January 2018, the story suddenly came alive again when the Wall Street Journal published a series of reports on how Trump’s attorney, Michael Cohen, made the $130,000 settlement payment to Stormy Daniels’s attorney.

The following is a timeline of known events surrounding the Stormy settlement, compiled from a dozen or so different reports that have come out since the WSJ story with different details on what happened.

Sept. 30, 2016
Cohen forms Resolution Consultants, LLC, a Delaware entity. For unknown (and possibly ill-advised) reasons, Cohen chooses a name that is completely transparent about the LLC being created in connection with some kind of settlement.

Oct. 1-Oct. 11, 2016
A settlement agreement is negotiated and agreed to in principle, but never executed, because the $130K payment is not made to Stormy’s attorney as agreed.

An earlier draft of the Side Letter is leaked by Stormy to Jacob Weisberg over at Slate. It appears to specify that payment is to be made to “RCI” (possibly Resolution Consultants), and also contains what may be typos. For instance, in paragraph 3, the Side Letter mistakenly refers to “DAVID DELUCIA,” even though Trump’s pseudonym for this agreement is “DAVID DENNISON.” (Though not directly relevant here, it’s worth noting that one possible explanation is that Stormy’s NDA/Side Letter is a standard agreement Trump uses regularly, and “DAVID DELUCIA” was the result of someone’s failure to correct update the template to match the terms being used in the Stormy agreement.)

Oct. 12, 2016
On October 12, 2016, Stormy’s attorney writes a brief, two-word email to Cohen: “We good?”

From context, this is likely Davidson checking in with Cohen to confirm that the settlement is actually going to proceed as planned. (And, although this is reading a bit between the lines, it also suggests perhaps there had been previous hitches in the negotiation, causing Davidson to feel the need to check in with Cohen and confirm that, this time, all systems really are go.)

Cohen responds quickly, stating, “Yes. It’s Yom Kippur so the office is for all purposes effectively closed. I am in tomorrow but can speak for the next 3 hours via cell if necessary.”

Although somewhat ambiguous, the impression given by these emails is that:
(1) Stormy Daniels’ attorney was asking about the status of the payment to Stormy,
(2) Cohen was trying to reassure Stormy’s attorney that the payment would be forthcoming, but that due to the holiday and lack of personnel at the office, he might be unable to finalize the practical side of the arrangements that day.

Oct. 13-Oct. 16, 2016
No payment to Stormy is made. Stormy seems to lose patience and/or hope in Trump’s intentions to ever pay, so she decides to go public with her story. She appears to have been in discussion with numerous outlets (Slate, Good Morning America, and Fox all appear to have been working the story), and the Trump campaign may have been fielding questions from these outlets seeking comment on the story. (One outlet, The Smoking Gun, will in fact go ahead and publish a story on Oct. 18th).

Oct. 17, 2016
Stormy has given up on Trump making good on the settlement. According to the Washington Post, on the morning of the 17th, Stormy’s attorney emails Cohen twice, to inform him he must pay up immediately or the deal is off.

In an Oct. 17 email, an attorney for Daniels — a porn star whose real name is Stephanie Clifford — threatened to cancel the nondisclosure agreement by the end of the day. . . . A second email to Cohen, a short time after the first, said Daniels was calling the deal off. “Please be advised that my client deems her settlement agreement canceled and void,” Daniels’s lawyer, Keith Davidson, wrote in the email, which The Washington Post obtained.

At 10:21am that same day, Michael Cohen forms a new Delaware LLC. This one is called Essential Consultants, LLC. Two minutes later, at 10:23am, the previously created LLC, Resolution Consultants, LLC, is dissolved. (One might speculate that the name “Essential Consultants” was deemed a less conspicuous alternative to the original entity.)

And then, at some point on the same day that Stormy threatens to call of the settlement, the Trump campaign makes four payments to Trump-owned businesses. Although, as usual, there is a regular stream of disbursements from the campaign to the Trump organization during this time period, these four payments, when combined with a fifth and final disbursement to a Trump-owned business that was made on October 25th, total $129,999.72 – 28 cents shy of the exact total of the settlement payment owed to Stormy Daniels.

The four payments made on Oct. 17th are as follows:

Three of these payments (for $13,431.88, $79,043.94, and $18,731.90) are reported in Trump’s pre-general 2016 report filed on Oct. 27, 2016. One of the disbursements (for $18,731.90) is reported later, in Trump’s post-general 2017 filed on Dec. 8, 2016.

Oct. 25, 2016
The fifth and final of the disbursement in the series adding up to $129,999.72 is made by the Trump campaign.

Oct. 26, 2016
At 4:15pm, Cohen receives an email from First Republic Bank, confirming that “the funds” have been deposited into a specific checking account.
Cohen then forwards the email to his personal email account, and then forwards it from his personal account to Stormy’s attorney, without comment.

Oct. 27, 2016
According to the Washington Post article, additional email traffic between Cohen and Davidson takes place to confirm that the $130,000 payment has been successfully transmitted to Davidson’s trust account.

Oct. 28, 2016
Stormy signs the settlement agreement, as does Michael Cohen, though he signs as EC, LLC, both as its representative and its counsel.

In the copies of the settlement and the Side Letter that have been publicly released, there is no signature for David Dennison, a.k.a. Donald Trump.

So that’s the relevant background, which brings us back to my tweet on the Trump campaign disbursements. Based on the disbursements itemized in the Trump campaign’s finance reports, which you can obtain from the FEC’s website, in the eight day period between Oct. 17th, when Cohen was informed by Stormy’s attorney that the deal was off due to nonpayment of the $130K settlement, and Oct. 26th, when the bank emailed Cohen to confirm that the settlement funds had been deposited in an account, there were five disbursements from the Trump campaign to Trump-owned businesses that total $129,999.72, or almost the exact amount of the settlement payment that Cohen needed to come up with.

The Stormy Disbursements: A Possible Scenario

If these five disbursements – let’s call them “the Stormy disbursements” for simplicity’s sake – are in fact the source of the Stormy settlement payment, there needs to be an explanation of how that money moved from the Trump campaign to the trust account of Stormy’s attorney. Based on the above timeline, here is my interpretation of what may have occurred here, based on the currently known facts.

In the fall of 2016, Cohen and Davidson negotiated an agreement in principle to buy Stormy’s silence (as well as the copyright for the images Trump texted to Stormy… and no you probably shouldn’t think too much on what that might mean, or you’ll regret it). In exchange for a mutual release of claims, and a payment of $130,000 to be wired to Stormy’s attorney, Stormy would agree to keep all information falling under the general header of “Trump’s affair with a porn star” a secret, as well as to hand over all evidence documenting that affair.

In order to keep Trump’s involvement in the settlement (and the affair) as secret as possible, the settlement agreement was drafted as a sort of complicated shell game, in which the document purporting to be the settlement agreement would contain all the terms, but would not identify either Stormy or Trump as its parties. A side letter confirming their identities as parties to the agreement would then be executed, and locked away as attorneys’ eyes only material.

On September 30, 2016, Cohen created Resolution Consultants, LLC, with the apparent intention to use it as the vehicle through which the settlement payment could be transmitted to Stormy. By early October, a draft settlement agreement had also been prepared and circulated among the parties. At some point, Stormy took a photo of the initial draft of the Side Letter, and sent it to the reporter at Slate. That draft contained a few differences from the final version that was ultimately drafted, such as the (mistaken?) use of “DAVID DELUCIA” as a pseudonum for Trump, as well as specifying that payment was to be made through “RCI,” a.k.a. Resolution Consultants.

But at some point in mid-October, things began to fall apart. Perhaps a payment deadline was missed, perhaps Cohen was acting shadier than normal, but at any rate something tipped off Stormy’s attorney, Keith Davidson, who began to grow wary of whether Trump, via Cohen, was truly acting in good faith, and whether the settlement payment would ever be made. On October 12th, Davidson sent the “We good?” email to Cohen, to check in on the settlement’s status, and Cohen used the Yom Kippur excuse in his reply to explain why no payment had yet been made. This suggests that the Trump Organization itself was involved in some way, and Trump Organization action would be required to effectuate the payment – otherwise, Cohen’s attempt to use the holiday as an excuse makes less sense.

But it would quickly become apparently that Yom Kippur was not the actual cause of the settlement’s delay, because several days would go by and still no payment would be be made to Stormy’s attorney. At this point, it’s not clear why Cohen did not transmit the $130K as agreed. It could be that Cohen/Trump had never actually intended to pay Stormy at all – that their intent all along had been to stall and delay Stormy until after the election, at which point it would be too late for her story to make a difference. Or perhaps Cohen/Trump really had intended to carry out the deal, had wanted to make the $130K payment if they were able, but they were blocked in their attempts to make the payment by unknown hurdles.

Either way, Davidson was not impressed with the delay. Finally, on October 17, 2016, Stormy’s attorney threw down the ultimatum: get Stormy her money by the end of the day, or else Trump’s affair with the porn star is going to be top of the hour news on every cable news show in America.

So now Trump’s Mr. Fixer was desperate. There was no more room for stalling, no more room for delay. He needed to find a way to credibly reassure Davidson that payment would be forthcoming, and he needed to do so quickly. Based on Davidson’s emails, Cohen needed to find a solution before the close of business that day, or else it would be too late.

Most likely, there were further exchanges between Cohen and Davidson on the Oct. 17th. Cohen must have wrote Davidson an email, or had a call with him, and said something along the lines of, “Look, the money is coming, for real this time, I promise. Trump is good for the hush money, it’s just taken a bit longer than expected, that’s all.” But by this point, words would not have been enough – Davidson likely would have wanted some kind of tangible assurance that, yes, payment really was going to be made soon.

Given that Stormy did not go public, it would appear that, one way or another, Cohen succeeded that day in convincing Davidson not to call off the deal right then and there, which would have bought him (and Trump) a little more time. But only a little. One way or another, Cohen still needed to get the settlement money together fast. So here’s the $130K question: how exactly did Cohen get the money?

Here are some possible options Cohen had:

1. Pay the $130K himself. Ahahahahaha no. No. Never did Cohen intend to bear the cost of this payment himself, and he wasn’t even going to temporarily bear the cost either, if he could at all avoid it. (Cohen has recently claimed that he obtained the $130K through a home line of credit, but the only universe that could have occurred in was if it was a desperate attempt at a temporary bridge. Cohen’s self-serving statements in this respect should also not be taken at face value.)
2. Have Trump personally pay the $130K. Too risky. There is no way of paying the settlement out of Trump’s personal accounts that would allow Trump to plausibly deny any knowledge of the settlement negotiations.
3. Have the Trump Organization pay the $130K. Maybe, but this option carries its own risks. A Trump Org payment may not necessarily impute Trump’s direct knowledge of the transaction, but $130K isn’t the kind of rounding error a company is likely to miss. To explain the money going out, Cohen needs to explain first where the money was coming in.
4. Have the Trump Campaign pay the $130K. There are risks here too, but this might make more sense in some respects. After all, there is this big pool of money just sitting there, and being paid out in large amounts to all kinds of recipients every day. And the whole point of that money is to use in helping Trump get elected. And – let’s be real here – although proving intent in a court of law is another beast all together, as a practical matter, we can all agree that helping Trump get elected was the whole purpose of the Stormy settlement.

One available option to Cohen – and the option he may have chosen – was to use a combination of the Trump campaign and the Trump Organization in a way that effectively amounted to his own in-house money laundering operation. Because the campaign and the company are (in theory) two entirely separate entities, doing business at pseudo-arm’s length. The Trump campaign already had an established pattern of making disbursements to the Trump Organization at regular intervals, and those disbursements were for legitimate services and often fairly large in their total dollar amount. Hypothetically speaking, if you were a money launderer, this might seem like a pretty sweet set up. There’s no need to explain where the money is coming from, because the Trump campaign has plenty of money that was legitimately obtained through donations, and explaining where the money is going out is easy, because you also control the entity that is requesting and receiving the payments. You can have the Trump Organization charge the Trump campaign whatever price you feel like, give or take a 100% markup, and the Trump campaign isn’t going to complain about it. And, as an extra bonus just to make everything even better, the underlying transactions are for large-tickets items with an inherently amorphous real value. Sure, event services have a ballpark range in price, but if you charge $28K for an event, who’s to say that objectively there is no basis for charging anything more than $16K?

So in a perfect world, what Cohen should have done is divide up the $130K intended for Stormy, and then transmitted it from the campaign to the Trump Organization as a nearly invisible surcharge on top of the Trump Org’s existing invoices to the Trump campaign. So that $20,000 invoice the Trump Org is planning to issue to the campaign for hosting that Trump rally at Doral? Make it $40,000 instead. Heck, maybe make it $50,000, it’s not like anyone is going to question an extra $10K when it comes to luxury resort event hosting. You still want to try to keep it superficially plausible, but you have some flexibility there. So you keep on like that, adding on a bit of padding to the invoices from the Trump Org to the Trump campaign, until that padding totals the needed number – in this case, $130K.

But in order to transmit the $130K through inflated-but-otherwise-legitimate invoices, you first need to have some otherwise-legitimate-invoices to inflate.

And this is why the disbursements from the Trump campaign to the Trump Organization that began on October 17th look so significant. Because Cohen was not in a perfect world, and Cohen did not have the time it would take for enough legitimate invoices to roll through to cover all of the $130K. Stormy needed that money now, within days or perhaps a week at the very most, or else the whole game would be lost.

So Cohen was desperate, and quite possibly out of other options. Based on the previous delays in getting payments to Stormy, there’s a good chance Cohen had already explored other routes at that point for getting the money to Stormy, but had not been able to find a suitable way of doing so.

Which is why it is plausible that, on Oct. 17, 2016, Michael Cohen might have done something very, very stupid: created fake invoices for the Trump Organization to submit to the Trump campaign to cover the amount that Cohen needed to pay to Stormy. (For the record, it probably wouldn’t have even taken desperation for Cohen to have done something this stupid, and if you need further proof of this, then I would like to submit this Esquire article as Exhibit A.).

The available evidence is consistent with a scenario along these lines occurring. Because on October 17th, the Trump campaign made four large-ticket disbursements to the Trump Organization, and on October 25th the Trump campaign made another disbursement that just happened to round out the total to within 28 cents of $130,000. And then, on October 26th, Cohen suddenly came up with the $130,000 he needed to pay Stormy.

Just a Coincidence?

One reasonable question people have raised in response to my tweets about the Trump campaign disbursements is whether this might all just be some kind of crazy random coincidence. Perhaps this is the kind of thing where, although the result might look surprising on the surface, in reality the result is something you might statistically expect to find given the data being looked at. After all, the five disbursements that total $130K are not the only disbursements that were made by the Trump campaign to the Trump organization during the time of the negotiations between Stormy and Trump. There are other ways in which these disbursements can be combined, and different amounts they might add up to.

So, could the fact these five payments equal $130K just all be a coincidence? Some kind of meaningless statistical quirk?

I am not a stats person, and I am in no way qualified to give any sort of mathematical analysis of this question. Luckily, a lot of people who are qualified to do so have become interested in the question too, and they’ve come up with ways of modeling the probability of the “Stormy disbursements” being a coincidental result.

But before getting into that, I’d like to address one issue I’ve seen raised a few times now: that these disbursements are not significant because they occur alongside other, non-significant disbursements. This claim, if true, would essentially make it impossible for money laundering to ever be detected, because hiding a stream of money laundering payments buried in a larger stream of legitimate payments is kind of the whole point. In fact, it would probably be more surprising if there were no “normal” disbursements from the Trump campaign to the Trump Org during the 8-day period at issue here, because it was rare for more than a few days to go by without some kind of Trump campaign-Trump Org disbursement being made. Throughout all of 2015 and 2016, there was a steady stream of payments being made out to the Trump Organization for various items: hotel stays, event planning, event facilities, meeting expenses, restaurant services, rent for office space, rent for other kinds of real estate undefined in the campaign finance reports, and so on.

In the relevant date range here, from October 17, 2016 (the date the payment vehicle LLC was created) to October 25, 2016 (the day before Cohen got the confirmation email that the deposit had been made in what is presumed to be EC LLC’s bank account), there were a total of eight disbursements made by the Trump campaign to the Trump Organization. These disbursements are listed below, with the five disbursements that make up the $130K in blue, with the other three in yellow. (Note that the possible “Stormy disbursements” constitute every large Trump campaign-Trump Org disbursement that occurred on October 17th, and excludes only what appears to be a one-night hotel room fee for Trump’s New York hotel. There were no additional big-ticket items on the 17th that could’ve been used as the basis for an inflated invoice, if Cohen was trying to go that route.)

And here is where we get back to the math. Given that the $130K total in payments includes some but not all of the disbursements in that time period, what are the odds of the relevant disbursement records coincidentally matching the amount of the settlement paid to Stormy at that time?

The answer is “very bad.”

Will Stancil and his brother Benn Stancil (co-founder of Mode Analytics) took a crack at modeling the probability of the “Stormy disbursements” being the result of chance, and wrote up their findings in an article here:

To explore whether these payments are worth investigating further, we have approached the question from another angle. Instead of examining the individual payments forensically, we have instead focused on the rather close match between the summed payments ($129,999.72) and the $130,000 Daniels payoff. It struck us as fairly unlikely that, by chance alone, so few payments would sum to such a precise figure.

In order to investigate these suspicions, we developed 10,000 sets of simulated Trump campaign payments. Each set contained 10 randomly generated payments. We then searched each of those sets for the combination of payments with the total closest to $130,000.

The simulation confirmed that it is extremely unlikely that, by random chance alone, a set of payments near a specific date would almost equal $130,000.

For each of the 10,000 sets, we generated a “closeness” value — the difference between their “best match” and $130,000. For instance, if the “best match” was $130,014.29, the “closeness” value would be $14.29.

Across 10,000 sets of simulated payments, the 99.9th percentile of closeness was $0.24. The actual degree of closeness in the real-life Trump campaign finance records is $0.28. In other words, out of every one thousand simulated payment sets, only one contained a combination of payments as close to $130,000 as the real-life payments made in the week preceding Oct. 25th.

The Stancils also ran a more conservative model, which was more generous in assumptions about how large the data set might have been (i.e., to represent disbursements selected over a longer range of time). Even by making the available time period more generous, however, the odds of an amount close to the Stormy payout amount coincidentally appearing are still very low:

To ensure that slightly altering our assumptions would not dramatically change these findings, we also tested a slightly more conservative model, in which 10,000 sets of 15 payments were examined. To keep the model computationally feasible, we restricted the output to the closest combination of six payments generated by each set.

In this model, the actual observed combination of $129,999.72 falls somewhere between the 98th and 99th percentile of closeness. In other words, while the observed outcome is considerably more likely under these very generous assumptions than under the normal assumptions, it is still quite unlikely overall. The output of both models are in the tables below.

I also heard from Zack Dennis – whose experience in combinatorics comes in part from his work on keyboard replacement methods over at asetniop.com – also got curious about the probabilities involved here, and emailed me to summarize the calculations he’d run to predict the probability of the “Stormy disbursements,” based on Trump Campaign to Trump Organization disbursements 2015-2016. He took a slightly different approach than the Stancils did (who were a bit more generous in their underlying assumptions), but got a similar result.

Zack examined the question in two ways. First, given all the disbursements from Trump campaign to Trump Org, how many random combinations of 8 disbursements contain a set which can add up to $130K? And second, given the chronological list of disbursements, how many times did the Trump campaign make 12 chronological disbursements to the Trump organization, from which 12 disbursement some combination of disbursements can add up to $130K?

For the first question, here is how Zack explained his model to me:

Imagine you have a data set of four numbers [1,2,3,4].  How many different ways can these numbers combine as combinations of two or more, where order doesn’t matter?

[1,2] = 3

[1,3] = 4

[1,4] = 5

[2,3] = 5

[2,4] = 6

[3,4] = 7

[1,2,3] = 6

[1,2,4] = 7

[1,3,4] = 8

[2,3,4] = 9

[1,2,3,4] = 10

That’s a total of eleven different combinations. Let’s say you’re interested in all the possible ways these numbers might add up to 7.  There are two possible combinations: [3,4] and [1,2,4].  That’s two out of eleven.

Now imagine instead of using the original data set of [1,2,3,4] we use the *entire* set of entries in the FEC spreadsheet that you posted starting from the beginning: [$1380.54, $9583.33, $37993.04, $3240.96, $9583.33…] and perform the same process:

[$1380.54, $9583.33] = $10963.87

[$1380.54, $37993.04] = $39373.58

…and so on.

What we’re looking for here are combinations that add up to totals close to $130,000.  In order to simplify the calculation process, I added a few conditions:

1. I used a total range of one dollar – from $129,999.50 to $130,000.50 – as a target. Your original calculation was within that range at $129,999.72.

2. I removed all amounts that were exact multiples of $1000.00 – $15k, $9k, $6k, etc. It would be too easy to “cherry pick” a series of these numbers to hit exactly $130,000.00, so we’re not going to consider them (plus, none of the numbers in your original data set were exact multiples – we’re trying to replicate the odds of what you did happening by chance). Plus, it saves some time computationally.

3. I considered up to eight separate numbers per combination (i.e. [a,b,c,d,e,f,g,h] = total). Building bigger subsets would take too long to process – I wrote my code in javascript, which is not particularly fast, and it took around 12 hours to process.

The total number of possible combinations was 172,325,161,239 (that’s 172 trillion).

The total number of combinations that fit the $130k range was 933,281.

That works out to a percentage of 0.00054%

So based on those numbers, the odds of *randomly* finding any batch of up to eight entries in the spreadsheet that add up to a number that’s approximately as close to your calculation are 1 in 18,464.

Now, the first question Zack was looking at considered combinations of any 8 disbursements, regardless of when they were made during the campaign. So each set could include disbursements made anywhere from April 2015 to December 2016. Which, as you might have realized, isn’t entirely analogous to the question at issue here with the “Stormy disbursements,” because any money laundering transactions of the sort I’ve theorized about here would have occurred during a defined window in time, not randomly throughout the entire campaign. So Zack also turned to look at a second question: using only chronological sets of disbursements (i.e., sets of 12 disbursements that occurred one after another), how many sets of chronological disbursements are there which contain disbursements adding up to within a dollar of $130K?

The answer: exactly one. The only disbursements that fit this criteria are the five “Stormy disbursements.” And those disbursements just happen to have taken place within the exact 9-day period during which any theoretical money laundering would have had to occur.

The Possible Origins of the Stormy Disbursements

From a mathematical point of view, the odds of the “Stormy disbursements” occurring by chance are very, very low. By my reckoning, that should be sufficient on its own to warrant further questioning here. But it also does not prove that any misconduct occurred, because although the odds of this happening by chance are pretty low, by definition even low probability events have to happen once in a while. So the statistical models alone can’t rule out the possibility that the coincidence of the “Stormy disbursements” matching the Stormy settlement payment is just a result of winning the bad luck lottery here.

Which means we need to look at what we know about the Stormy disbursements beyond their simple dollar amounts. While any conclusive answer here would require access to financial records that would be impossible for me to obtain, what I can do is evaluate whether, based on existing public records, any of the “Stormy disbursements” can be shown to be legitimate campaign disbursements. In other words, whether any of those five disbursements totaling $130K can be convincingly explained away as real expenses paid by the Trump campaign for real services provided by the Trump Organization.

And we at least have an idea of what to look for here to try and show that these transactions were legitimate. Because if the “Stormy disbursements” were in fact merely a vehicle used to transfer $130K from the Trump campaign to Trump Org, then they must be either one of two things: (1) payments for fictitious services that were never rendered to the Trump campaign, or (2) duplicate payments for real services that the Trump campaign received, but should not have been paid for twice. In other words: are there real expenses that can be tied to the Stormy disbursements, and is the reason to thing the Stormy disbursements were not duplicate payments for those expenses?

Unfortunately, because of the limited data available from FEC reports, we’re not able to conclusively identify the events that any Trump campaign-Trump Org disbursement may have been associated with, but based on the larger pattern of disbursements from the Trump campaign to Trump Org, there are reasons to suspect that the Stormy disbursements may lack a credible explanation.

First though, a couple caveats about the limits of using the FEC records in an attempt to correlate campaign disbursements with actual campaign events:

• Not all Trump campaign events at Trump-owned properties have associated campaign disbursements, suggesting either that on at least some occasions the Trump Organization did not charge the Trump campaign for events, or that payment for these events was not identified obviously in the disbursement records for whatever reason.
• Matching up a particular Trump campaign disbursements with a particular Trump campaign event at a Trump property will not always be possible, as there is no definitive list of “Trump events” to compare against, and it’s theoretically possible that an otherwise inexplicable disbursement is associated with some private Trump event that was not publicized enough to leave a record.

This all makes analyzing the nature of these disbursements difficult, and prevents me from reaching any firm conclusions here based only on the FEC records. Though, on the other hand, it is worth noting that, from a money launderer’s perspective, this inherent ambiguity in the data is something of a bonus. Yes, the campaign disclosure requirements do remove an element of secrecy, but not enough of one on their own to allow for any conclusions that money laundering occurred.

Given these caveats, then, what can we conclude here about the five “Stormy disbursements” based on publicly available information?

The Trump National Golf Club Washington DC Disbursement ($8,544.00)

Let’s start with the Oct. 17th disbursement of $8,544.00 to the Trump National Golf Club Washington DC, out near Sterling, Virginia. This is probably the easiest one to start with, because Trump’s DC golf club was not a hot spot on the campaign trail, which means there are only a few transactions for us to contend with. In fact, I could only find a single Trump campaign event that took place there: an October 25th event that was headlined by Trump’s daughter-in-law, Lara Trump.

The event was a relatively small one, and was hosted by “Diverse Communities in Virginia Supporting Trump.”

It was supposed to have been preceded earlier in the day by a campaign rally in Loudoun County, but the event was cancelled as Lara “was stuck in traffic” and could not make it. She did make the Trump National Golf Club DC event, though, and in her remarks to the group, the future president’s daughter-in-law urged those present to disregard anything the media said about Trump: “Don’t believe anything the media tells you. They’re all liars. And now not only do we know they’re liars, we know that they’ve been in Hillary Clinton’s back pocket the whole time.”

Which means there was a genuine event that took place at the Trump National Golf Club DC, and that event could provide a possible explanation for the source of the $8,544.00 disbursement. Although the Trump campaign’s regular practice was to pay the Trump Organization after an event at a Trump property, that does not mean there could not have been occasions on which the Trump campaign made disbursements to the Trump Org in advance.

But there’s a catch: the $8,544.00 disbursement is not the only disbursement that was made to Trump National Golf Club DC in October of 2016. Here is a list of all the disbursements that were made to Trump’s DC golf club in the 2015-2016 campaign (with the possible “Stormy disbursement” marked in blue):

There were two large disbursements made to the club in all: one for $8,544 on Oct. 17th, and one for $11,600 a couple weeks later, on Oct. 31st. Both payments are within the ballpark for what it would cost someone to host a Tuesday all-evening wedding with full buffet dinner at the Trump National Golf Club DC, which means either could have been issued in connection with Lara’s event (though, her event was only two hours and just hors d’oeuvres were served, so the Trump campaign’s payments would seem to have been a bit generous). Which means if the $8,544 payment was a prepayment to Lara Trump’s event, there must necessarily have been two campaign events that took place at Trump’s DC golf club in the weeks before the election, both of which cost approximately $10,000. I have yet to find evidence of a second event that could explain these charges.

Trump International Hotel Washington DC Disbursement ($13,431.88)

Next up is the Oct. 17th disbursement of $13,431.88 to Trump International Hotel in Washington, DC, a.k.a. the Old Post Office building. In the FEC reports, this disbursement is identified as payments for “FACILITY RENTAL/CATERING SERVICES.”

This payment is tricky to nail down because of the sheer volume of activity in DC, and at Trump’s DC hotel, that took place during the campaign. Additionally, there are more reports of campaign events and appearances at the Trump International Hotel than there are disbursements to it, suggesting that not all campaign activity at the hotel was compensated by the campaign.

Additionally, the campaign at times appears to have made indirect disbursements for events at Trump properties. For example, the campaign did not appear to make any direct payments to the Trump International Hotel in DC in connection with the national security meeting hosted there on March 31, 2016, but the Managing Director for Trump’s DC hotel and the General Manager for Trump’s Chicago hotel both appeared to receive payments directly from the campaign in connection with the event:

Which means any kind of direct 1:1 matching between campaign events and campaign disbursements for the DC hotel is probably not possible based on available data. However, if we look only at disbursements made directly to the hotel itself, in total, there were seven disbursements during the campaign (with the possible “Stormy disbursement” marked in blue):

That first payment on January 12th can be disregarded for the moment, since it could not be connected to any events taking place in the weeks before the election. Which means, based on the campaign’s known events, the best possible explanations for the October 17th disbursement of $13,431.88 is likely in connection with one of two possible events:

• The Sept. 16, 2016 “birther” press conference/hotel informercial; or
• The Oct. 26, 2016 hotel event that Trump appeared at

As an initial matter, if the $13,431.88 was in fact in connection with the Oct. 26th event, then that might mean the payment wasn’t connected with Stormy, but the Trump campaign would still have a problem here. Because that event was not a campaign event – it was a ribbon cutting/grand opening ceremony for the hotel – and Trump was in attendance not as a candidate but as the business owner. As such. campaign funds should not have been used in connection with the event. (Notably, this would also apply to the $956.08 lodging charge on Oct. 27th, which looks suspiciously like a charge for someone’s hotel room for the night following the grand opening ceremony…)

The Sept. 16, 2016 Press Conference at Trump’s DC Hotel

So excluding the Oct. 26th ribbon cutting, that would leave the infamous “birther” press conference on Sept. 16, 2016 as a possible explanation for the Oct. 17th “facility rental/catering services” disbursement in the amount of $13,431.88. But once again, as with the Trump National Golf Club DC, we’ve got a problem: there are later disbursements for similar values that can’t be explained by known campaign events. Two of them, in fact. One on Nov. 2nd, for $9,245.93, and one on Nov. 14th, for $12,265.61. I think it’s fair to assume that one of these three $9K+ payments was for the Sept. 16th event, but there is no way to determine which one. Nor is there a ready explanation for what the remaining two would have been for. Although it’s possible the Oct. 17th disbursement was legitimate, it could equally be a duplicate payment for either the Nov. 2nd or Nov. 14th disbursements, and based on available data there would be no way to tell.

Trump International Hotel & Tower New York Disbursement ($10,248.00)

This disbursement was the final one in the “Stormy disbursements” series, and the only one of the five disbursements to have occurred on Oct. 25th.

There is also not all that much that can be said about this disbursement, because there just isn’t much to go on in terms of linking it to any known Trump campaign activities. The charge is, presumably, for people affiliated with the Trump campaign for their lodging at Trump International Hotel & Tower. This was a frequent occurrence during the campaign, so obviously something that happened with regularity (at least from August to November), and based on available data, there is no apparent way to evaluate its legitimacy. However, the Oct. 25th disbursement does stand out in one respect – it is the single largest disbursement (below, in blue) that the campaign ever made to Trump’s New York Hotel.

So once again, whether this disbursement was legitimate or not is not something that can be determined based on available data. Invoices for all Trump International Hotel & Tower disbursements would be required to ascertain whether there is in fact an underlying campaign charge associated with it.

Trump International Hotel Las Vegas Disbursements ($18,731.90 & $79,043.94)

Finally, that brings us to the fourth and fifth – and most suspicious – of the “Stormy disbursements,” and that’s the two disbursements to Trump’s Vegas hotel for $18,731.90 and $79,043.94. Both disbursements took place on Oct. 17th (though only the $79K disbursement would be correctly reported in Trump’s October 27th FEC campaign finance report; the $18K disbursement would not be reported until Trump’s December 8, 2016 filing). Here all the disbursements made by the Trump campaign to Trump’s Vegas hotel, with the two “Stormy disbursements” marked in blue.

These two disbursements are notable for a number of reasons, but perhaps the most obvious one is the sheer dollar figure involved. The combined total is $97,775.84, but even standing by itself, the $79,043.94 disbursements was at that time the single largest disbursement the Trump campaign made to a Trump hotel. There were larger disbursement made to hotels for lodging during the course of the campaign, but only if you count the campaign’s July four-day stay at the Cleveland Westin during the Republican Convention in July, and the election night event at the Manhattan Hilton. There was also a campaign rally that the Trump campaign decided to throw last minute after the first debate — and which the campaign didn’t even organize until the morning of, which on the way to Hofstra University for the debate — and the results of that last minute scramble cost the campaign $97,419.95. So yes, these things can add up, and lodging costs are a significant expense for campaigns. But were they significant enough to explain the $98K disbursements on October 17th?

There are two possible events that might explain the Oct. 17th charges:

• The third presidential debate in Las Vegas on Oct. 19th
• A Vegas campaign rally on Oct. 30th

The Oct. 30th rally is a very poor fit for the Oct. 17th disbursement, and I do not believe it provides a viable explanation for those expenses. First of all, the Oct. 30th rally wasn’t even at Trump’s hotel. It was at the Venetian, owned by Trump-supporter Sheldon Adelson. Nor were there any follow-up events in Vegas that might explain the additional charges to Trump’s hotel, because after the 11am rally at the Venetian, Trump left town for two other rallies that day, in Colorado and New Mexico. And we know that after the Oct. 30th rally, the campaign paid the Venetian (or rather its sister hotel) for the rally – to the tune of $114,246.06.

So a total of $214K, for a rally that lasted maybe a couple hours? That seems extreme even by Trump’s standards, and not a particularly plausible explanation. Which leaves the other possibility: that the Oct. 17th charges were prepayment for the Oct. 19th debate in Vegas. After all, Trump did stay at his hotel while there — something he even brought up in the debate. Although the $100K price tag was described as “lodging” on the FEC reports, this may have been in error. Although lodging alone could not possibly account for the size of the disbursement, a campaign event plus housing a few dozen staffers in the hotel’s priciest rooms could do it.

And there was a campaign event the night of the debate. Well. Sort of. There was this, anyway:

Although, according to reports from attendees, the whole things sounds like it was a slightly sad affair.

 “We’ve got Diamond and Silk, what do you think Hillary’s got?” Boris Epshteyn asked me. “Come on, admit it, Republicans are more fun!”

Epshteyn is a senior adviser to Donald Trump’s presidential campaign and late Wednesday evening, after the debate, he stood at the head of a long table in the half-empty restaurant of the Trump International Hotel here, drinking and eating jovially with the rest of the staff. …

Next to Epshteyn was Kellyanne Conway, the campaign manager who sipped red wine and ate carrot sticks, and next to her Rebekah Mercer, the deep-pocketed Republican donor who looks like she could be related to Sarah Palin (though Palin was in the “spin room” after the debate, she wasn’t at the party). The two women whispered and hugged.

While $100K seems like a hefty sum for such a sparsely attended event at Trump’s own hotel restaurant, I can’t say for certain that it would not have cost that much. (Although I will say definitively that if it did cost that much, Trump was ripping himself off here.) Still, this possible explanation runs into the same sort of problems as the other disbursements. First, it would be a prepayment, made in advance of the event itself, which just is not the typical way the Trump campaign operated with regard to Trump Org expenses. And second, although the debate might explain the Oct. 17th payments, that leaves us with more disbursements to the Trump International Hotel Las Vegas than there are events to explain them, which in turn leaves us with the question of how to explain all the other disbursements to Trump’s Vegas hotel.

Because following both the Oct. 19th debate and the Oct. 30th rally at the Venetian, there were, in all, a total of $174,343.58 in additional disbursements to Trump’s Vegas hotel. In fact, one of those disbursements – an Oct. 25th disbursement for lodging, for $16K – would, on the surface, appear much more consistent with usual campaign lodging expenses. But if the lodging expense was not for the Oct. 19th debate, what was it for? And if the Oct. 17th payments really were prepayments for either the Oct. 19th debate or the Oct. 30th rally, what on earth then could be the source of that $150K in total disbursements in late November?

And wait, there’s another complication here. The Trump campaign is not the only entity spending money at Trump properties. In fact, for the Vegas hotel, there were four other disbursements in October and November – three from the Trump Victory PAC and one from the RNC:

That means we have an additional $67,545.30 in spending at Trump’s Vegas hotel being disbursed by Trump-supporting causes in the weeks before the election. And while perhaps I have no conclusive evidence to refute it, based on the public reports of Trump’s campaign activities that I’ve been able to find, I just don’t think I am going to be able to be convinced that the Trump events in Las Vegas in the 2.5 week run up to the election are sufficient to explain the total of $347,074.95 in disbursements to Trump’s business – and that amount is before you even get to any payments associated with the actual rallies he had there!

All of which is a long way of saying this: no, the available records cannot prove that there was no legitimate basis for the “Stormy disbursements,” but neither can they provide any basis from which we can conclude they were in fact legitimate expenses. Given the statistical unlikeliness of the “Stormy disbursements,” the lack of a documented explanation for those expenses means further investigation is warranted.

The only way to verify whether these disbursements were issued as compensation for genuine campaign costs would be to review all the underlying invoices and payment records, to confirm both (1) the invoices exist for an actual service provided by the Trump Organization, and (2) the Oct. 17-25th were not duplicate payments for real Trump campaign events that had already been (or would be) paid by the Trump campaign under a different invoice. And for that to happen, either both the Trump campaign and the Trump Organization will have to voluntarily release records to verify these transactions, or else someone with the ability to compel production will have to decide to seek them.

The Trump Organization’s Controls

Even aside from the “Stormy disbursements,” there is a bigger story here that, to date, I do not believe has been adequately explored or reported on. From a compliance perspective, the Trump campaign’s ongoing use of the Trump Organization as a vendor creates a serious corruption risk, and I’ve seen no indication of controls in place, in either the campaign or the Trump Organization, that could have adequately mitigated the inherent risk created by the commingling of the two organizations. That inherent risk, when combined with the Trump Organization’s lengthy record of eschewing traditional notions of corporate compliance in both its domestic and foreign operations, creates the perfect conditions for misdirection of funds to occur.

That the Trump campaign and Trump Organization did not have sufficient controls in place can be demonstrated here by a few examples which, presumably, did not involve corrupt intent, but which nevertheless show that the way Trump ran his campaign made it possible for the unwarranted event charges to be approved and go undetected for months on end. To start with, let’s take Trump’s Doral property. Trump had lots of events there over the course of the campaign, and disbursements of greater than $1,000 were common. And, at least once, the Trump campaign disbursed over $10,000 to Trump’s Doral business for services that were never rendered.

To show what happened, the following is a list of all disbursements made to Doral from that Oct. 28th through the end of 2016:

In this case, the problem with the disbursements is fairly obvious, because two of the payments are for the identical amount of $11,541.20, an obvious sign that there could have been a double payment. And yet, it would not be until four months later that someone discovered the error. Here are all of the records for the disbursement at issue:

We don’t know exactly what happened here, but clearly at some point in Nov. 2016, someone signed off on a payment to Trump National Doral that was not owed and should not have been paid, and the controls in place were insufficient to stop the transaction or detect the problem. But to show what a problem this situation is, just imagine what would have happened had the second payment to Doral not been for the exact amount as the previous payment, but had instead been for, say, $16,841.96. A totally plausible amount based on disbursement patterns. And a totally plausible expense, given the apparently frequent campaign events that took place at Doral. Based on FEC records, it would be impossible to conclude one way or another if the expense was legitimate. And, moreover, it is impossible to know if the duplicate payment would ever have been detected, without the tip-off of the amounts matching exactly.

And that’s not the only time in the last months of the campaign that such a mistake occurred. Heck, the campaign twice accidentally overcharged itself for hundreds of dollars’ worth of Trump-brand water bottles that later had to be reversed:

Or take an example from another Nevada campaign event, this time in Reno, at the Sparks Convention Center:

Here, the mistakes were fairly obvious from the FEC filings, and were later corrected. But take the Sparks Convention Center example; what it does show is that a $27K check could be sent out from the Trump campaign when it should not have been, and then go undetected for over four months.

The wrongful disbursements discussed above appear to have been accidental, perhaps the result of some paperwork snafu, but this pattern strongly suggests that it would have been possible for someone in the Trump campaign with access to the pocketbook to have intentionally disbursed campaign funds for campaign events that never happened. Moreover, we also have reason to believe that someone disbursing campaign funds as payment for non-existent campaign expenses could have been able to do so without being immediately detected – because in the accidental disbursements discussed above, in which the duplicate payment were fairly obvious just from the FEC reports, the errors still went undetected until long after the campaign was over and Trump was well into his presidency.

Moreover, the question of whether any unwarranted disbursement might have been made intentionally – of whether the Trump Organization might have knowingly engaged in money laundering with respect to Trump campaign funds – is far from some kind of inflammatory or scurrilous accusation. Trump Organization’s connection to money laundering is a proven fact, many times over, and that’s even before you start looking at the real estate deals in Azerbaijan, Vancouver, SoHo, Panama, or a dozen other locations. I have no hesitation in stating that the Trump Organization is very familiar with money laundering, and has a long history of associating itself with experienced money launderers. (And if you doubt me, just go ask the United States Attorney for the Eastern District of New York just how kind Mr. Sater has been to share with them his wealth of knowledge on the subject.)

As for whether money laundering occurred here, I could not disprove beyond a reasonable doubt that some valid explanation for these strange and unlikely payment patterns might exist. Maybe Trump’s campaign operations were wholly legit, and the five “Stormy disbursements” is no more than the false positive equivalent of winning the lottery. Maybe Sam Nunberg was just making things up when he said on one of his many cable news appearances that there had been someone on the campaign who was stealing from it. Maybe there is nothing going on here.

But there are too many red flags present for the legitimacy of these campaign disbursement to be assumed without evidence. I’m willing to be wrong about the five “Stormy disbursements” being suspicious, but I’m not willing to simply assume that a crazy coincidence is the likeliest explanation here, when there exists a series of disbursements that look exactly like what you’d expect to find if someone had, in an act of desperation, engaged in financial shenanigans on Oct. 17, 2016 in a last-ditch effort to come up with $130K so that the porn star Trump had an affair with and who was threatening to go public could be paid off.

Luckily, I suspect Stormy Daniels is going to be all over this, should her suit remain in open court long enough for her to have the chance to look into exactly where Cohen got that $130K from. Which means perhaps we’ll get an answer here after all. So fingers crossed, and best of luck to you, Stormy.

-Susan