Calculating The Deficit: 4 Methods

For one untrained in the art of Deception Accounting, it would be easy to conclde that the deficit is a pretty straightforward concept…Revenue – Outlays = Surplus/(Deficit)…right? I thought so too for a while….taking the government’s reported Deficit # from the Monthly Treasury Statement (MTS) at face value. I even took the time to tie it out as I was gathering the data before I ever went live with this blog….comparing FY 2011 and FY2012 and finding only a $7B and $3B difference…good enough for government work right??

As it turns out…no. Over the last year, I have determined that those two data points are actually an anomaly, and that the reported deficit is quite different from a handful of other deficit calculation methodologies…including the methodology I use here when calculating the daily “cash” deficit. For example…per the MTS…the deficit over the last 12 months…May-2013-April-2014 was $500.097B. Compare that to the USDD methodology…at $569,646…nearly $70B higher.

2014-06-04 D1 VS D2

Basically….what this chart is saying is that the government is telling you the deficit over the past 12 months is $500B….and I’m telling you that it’s at least $570B. Is this a problem, and which number should we believe? To delve into that, let’s take a step back and first define four different ways one might go about calculating the deficit.


D1 takes the numbers straight from the MTS…this is the official deficit number. How it is calculated is anyone’s guess…my guess is that it’s loosely based on cash, but also includes accruals, and has cherry picked segments of government completely carved out. One large example of this would be the US post office, which operates under the farce that it is somehow a separate organization…and therefore USPS losses( and profits) are excluded from the deficit.

D2-Cash Flows:

This is the primary methodology I use to calculate the deficit based on the Daily Treasury Statement Cash Inflows and Outflows. It’s fairly straight foreward….in a given day, if cash deposits are $10B, and cash withdrawals are $15B…you have a $5B deficit. If your beginning cash balance was $50B….you would end the day with $45B.

Of course, it is never that simple, with the main complication that debt is continuously being issued and repaid…every single day, ranging from hundreds of millions to hundreds of billions on any given day. Fortunately….these are broken out on the DTS, and so we simply subtract them out conceding that issuance and repayment of debt do not contribute to the deficit directly. For example…going back to our simple example…say we also issued $100B of new debt and repaid $50B. Rather than ending the day with $45B cash…we would end up with $95B…but our deficit would be the same…$5B.

Finally, I pull out tax refunds…subtracting them both from revenues and outlays. The net effect of this on the deficit is zero, but I feel this is a more accurate accounting treatment…not making this adjustment would overstate revenue and outlays by about $350B per year.

Now…there is one flaw in this approach that I am aware of, and it relates primarily to debt issued at a discount. For example, just say we issue a $100 6 month T-Bill for $99…and pay it in full 6 months later for $100. In reality, what has happened is that we have is a $99 loan and $1 of interest. Unfortunately, this is not how the transaction is accounted for on the DTS…where the $100 out is all lumped into debt repayment…and therefore the $1 of interest is permanently excluded from the deficit in the D2-calc. In the big scheme of things, it’s not huge…$30-$50B understatement of the deficit annually, but it is a problem.

D3-Delta Cash Delta External Debt: (Key word…external debt only…D3 does not include the $5T of Intragovenmental Debt).

This is a very simple approach that in theory probably yields the most accurate deficit number available…in theory at least. Basically…I can ignore everything else and just observe the change in debt and cash between two periods to calculate the deficit/surplus over that period. Say I begin the period with $100B of cash and $10T of debt. If I end the period with the same cash and debt, my deficit….regardless of all the noise in the middle…was zero. We know this because mathematically….a deficit/surplus can only manifest itself in one of these places. So if we end the period with $10.2T of debt, and $150B of cash…we could calculate that we had a $150B deficit over the period.

There is one caveat however….this methodology breaks down during periods when the Treasury department uses “extraordinary measures” (EM) because essentially they just start pretending that they don’t owe money on some things….thus understating debt and completely nuking this calculation. We saw this back in October when the debt outstanding jumped $328B the day after the debt limit was raised (removed actually). So basically, any period that starts or stops within EM…D3 will be useless for calculating the deficit. However…once EM is stopped…all of the debt is brought back on the balance sheet and everything ties back out. So looking back to last summer, we had EM in effect between May and October….so D3 between say June and September would be screwed up…but I could look at April-November and get the right number.

In any case…I started this blog back in November 2012…right on the cusp of EM that started in December, and I had data from the prior period of EM from the summer of 2011. So while I hadn’t yet defined D3…I was aware that it would not work for what I was trying to accomplish….which is why I developed D2 and chose it as my primary metric….though admittedly I was not aware of the weakness discussed above.

D4-Delta Cash Delta All Debt: This is pretty much the same as  D3, but it expands and includes the full balance of Federal Debt. As discussed here, I’m not convinced this is a valid method since I don’t believe the $5T of intragovenmental holdings are real debt…rather they represent a cumulative accounting of funds already stolen/misused. Furthermore…I fundamentally believe that SS…whose “trust fund” represents more than half of that $5T balance is nothing more than a cover for a broad based ~15% income tax….disguised as a pension plan that is really a gigantic welfare scheme…as discussed in detail here.

So…I present D4 primarily as a reference point only, though I don’t completely dismiss its relevance.

And now, with D1-D4 defined, let’s line them all up…I chose the twelve month period of April-2013-March-2014 because it avoids last summer’s EM issues for D3 and D4 as noted above.

2014-06-04 D1-D4

So we have the official MTS deficit number at $494B…$145B lower than what I would call the best number..D3 at $639. Just for fun…let’s pencil that one out by pulling up the final DTS from March 2013 and March 2014. We see that over this 12 month period cash increased from $79B to $142B… a $63B gain. However, Debt held by the public increased from $11.917T to $12.619T…a $702B increase. So net the “bad guy” of $702B debt gain against the $63B “good guy” cash build….and we get a D3 deficit of $639B. You probably want to know what makes up the $145B delta…..that’s a damn good question…and one that I can’t answer. It has to be far more than just the post office being excluded…that was supposedly only $5B for FY 2013. But just because I can’t think of a good explanation doesn’t mean one doesn’t exist….any ideas from my readers?

2014-06-04 D1-D4 10 year

Above we take a final look…this time over the 10 year period from 4/2004 to 3/2014. Over that period, we see an $837B delta between the official deficit number and D3 ($2.9T if you believe in D4).

The question I asked at the beginning was

“Is this a problem, and which number should we believe?”

As I detailed above…D3 is the number I put the most faith in….cash and debt outstanding are separately verifiable and just about the last thing any organization is going to misstate…or worse lie about. I’m not saying they wouldn’t….just that they would probably lie about a lot of other things first….before resorting to lying about debt and cash, which are generally the easiest items for an auditor to verify. The MTS on the other hand….I’m not so sure. Give an organization some clever accountants and the power to define their own accounting policies and a standards and at the end of the day you will probably have some pretty useless financial statements….and that goes for corporate America too…not just Uncle Sam.

As for whether or not this is a problem… first thought was yes…. depending on what makes up that $837B 10 year delta between D1 and D3. If it is a systematic and pervasive intentional understatement of the United States deficit to the tune of nearly $1T…then I’d say yeah…it’s a problem…though at $1T…perhaps it’s barely material when bumped up against the big picture. It’s not so much the $ that are of concern…it would be the credibility of the government at stake….

But on second thought….since they don’t have any of that (credibility)….maybe it’s all just a big nothing either way….curious perhaps, but not as interesting as Dancing With The Stars Season 19???

Monthly Treasury Statement Vs. Daily Treasury Statement

I don’t think this will be a news flash for anyone, but as detailed here the Daily Treasury Statement (DTS) released daily by Treasury is the primary source of data used for all of the US Daily Cash Deficit posts. However, there is another Treasury report, the Monthly Treasury Statement (MTS), that the rest of the world is much more familiar with. Any time you hear an “official” revenue, outlay, or deficit number….odds are it came from the  MTS.

Now, you might be tempted to assume that the MTS is just the sum of the DTS…rolled up into a nice government report. Unfortunately, this simply is not the case. On several occasions, I spent more hours than I care to admit trying to tie the two out, failing miserably each time. After a chat with Treasury staff, I found out why…they are generated from completely independent data sets and simply are not reconcilable. During my attempts…I came across many many examples where the numbers from each report…in categories you would expect to be straight forward like NASA expenditures, were in fact off by 50-70% or more. As a result, as I’ve said before, I don’t like the MTS and I don’t trust the MTS.

Unfortunately, when I started blogging about the daily deficit last year, I made the assumption that for all it’s flaws, the MTS and the DTS were close enough. I based this on a comparison of FY 2011 and FY 2012, which over the twelve month period, despite monthly deltas, on the year was only off by $7B in 2011 and $3B in 2012…close enough for government work right?

Well, a few months ago, I decided to pull the historical MTS data and line it up with the DTS data, and while sure enough the 2011/12 FY are pretty darn close, not much else is. In fact, the average TTM difference over the last 5 years is $79B. At present, the TTM delta is $95B, with the cash deficit being $95B higher than the reported MTS deficit. Lesson learned…2 is not an adequate sample size.  This is likely to impact my world famous CBO vs CBO challenge, but I’ll tackle that another day.

While I can’t tie out the MTS to the DTS (since it’s not possible), I do have a pretty good idea of where at least some of the variance lies. The first thing you will notice about these two series is that DTS revenues are almost always higher than the MTS. The reason for this is that it represents the cash flows in for the entire federal government….and then some. My theory is that some of what ends up on the DTS is ultimately excluded from the MTS with the most prominent example being the US Postal Service. It would seem that USPS runs it’s business through treasury bank accounts, resulting in cash source of about $87B per year. On the outlays side….they have gone out of their way to make it difficult to track. They split out postal service money orders, but apparently not op costs. I assume that payroll gets grouped together with “Federal Salaries”. So as far as I am concerned, I can probably assume that the $87B of cash in is offset by probably $100B or so of outlays, contributing to about $13B of deficit (lot of assumptions there) However, on the MTS, you can probably bet that the revenues and outlays are completely excluded, and somewhere treasury makes an entry indicating a $13B loan…that I’m pretty sure will never be repaid…thus excluding it from the official deficit calculations. I catch it in the cash deficit…they purposely exclude it.

This is just the largest item….I’m sure there are more, plus untold amounts of “modified accrual” shadiness going on that account for the differences. So, I’ll just go ahead and say it….in my opinion the MTS is a bogus report that simply can’t be trusted. When all else fails….all we can rely on is the cash flow statement, and the DTS is the only timely reliable source of that for the time being. As long as government accountants can slice and dice and exclude whatever they want, you simply can’t trust the reports they generate.

So…why do I trust the DTS??? you probably are wondering? Well, I don’t dismiss the possibility for shadiness, but for the most part, the DTS is verifiable. Cash in less cash out damn well better be equal to delta cash and delta debt. Debt is separately verifiable, and cash is just about the last thing you want to lie about because it is the first thing any auditor is going to look for. Finally…these reports are generated daily with a 24 hour turnaround. Creating fraudulent financial statements takes time (I assume 🙂 ) , so the risk here is less than something bureaucrats have a few weeks to fiddle around with.



“Data Shift to Lift US Economy by 3%”

I thought this was funny. Apparently, the Bureau of Economic Analysis (BEA) is changing the accounting methodology for GDP to add a few new categories that will have the effect of increasing GDP by a few %. Here is the money quote:

“…so we are essentially rewriting economic history” – Brent Moulton

Isn’t that hilarious? Look…All you really need to know about GDP is that it is a completely bogus number…and yet…the pursuit of GDP is the dominant focus of nearly all economic policy. This error is responsible for a huge misallocation of global resources…but I digress.
In 1982, nominal US GDP was $3.3T…30 years later in 2012 it was $15.7T almost 5 times higher. So…think back to 1982…are you 5X better off? Are you any better off? Is the nation any better off? Sure…some of that is inflation, and population growth, but most of it is something else…Dark GDP being converted into something economists can kinda-sorta pretend to count.
Perhaps the best example of this is has been “housewives” leaving the home and entering the workforce. Imagine a community where 100% of the women are housewives. They perform the role of housekeeper, caretaker, chef, chauffeur, gardener, nurse, and psychiatrist…for their own families. Then…due to the changing times… all of the women in our hypothetical community decide to join the workforce and all find jobs making wages equivalent to their husbands. The economist will go crazy!!…GDP has doubled…or more. The community is obviously twice as well off right? Well…not so fast.
The truth is, in our hypothetical economy, GDP hasn’t come close to doubling. All that has really happened is that all of the work that had been done by the housewives off the books in the shadows of Dark GDP…has now come into the light to be counted by the economist. And while the working woman now has a wage, she now has to pay someone to do all of the things she used to do. So, when it is all said and done…I would grant that yes…GDP has definitely improved some because there are bona fide efficiencies to be gained in the outsourcing of the housewives duties. However, were true accounting even possible, I think it would show that around 80% of the gain was simply due to finally being able to count (and tax) the activities that were already going on. This is the true, and mostly untold story of the supposed”economic growth” we have seen over the last 50 years.
So my prediction.. only half joking is…that someday soon, the BEA will come up with a new change to GDP to somehow include in GDP not only the economic value of housewives, but also the value of the millions of people who are opting out of the labor force rather than be counted in the “unemployed” category. Currently, the rent an owner occupied home could have generated if the owner didn’t live in it is a part of GDP….so doesn’t it just  makes sense that we include the value of the services these people could be doing.. but instead have chosen not to? After all, it’s already a fairy tale number…why not give it a happy ending?

Calculating the Daily Deficit

When we talk deficit here at the Daily Deficit….we are talking about cash. Simply…cash in less cash out. If I take in $5B in taxes, and spend $10B…my daily deficit was $5B. We make two notable adjustments to the DTS totals. We start with total cash in, then we subtract cash in from issuing debt. This makes sense. Issuing $100B in new bonds increases cash, and debt by $100B. It has no effect on the deficit…it’s a pure balance sheet transaction. We do the same on the other side of the equation by removing cash outflows used to pay down debt, or pay off expiring bonds. This gets us down to revenue in from taxes and other sources, less cash out from expenditures. We then make one final adjustment. The DTS shows tax refunds on the expenditure side. As an accountant, that makes me cringe a little bit. I believe that tax refunds should be accounted for as a reduction in revenue, not an expense, so we take daily tax refunds and reduce cost by that amount, as well as revenue. If we don’t do this, we would essentially overstate revenue and cost by about $400B per year.

So, lets take a look at 11/8 and pencil out a quick example. Per the DTS:

Total Deposits   ($131.251B) – Debt Issues ($127.671B) – Tax Refunds ($0.131B) = $3.449B of Revenue

Total Withdrawals ($110.529B) – Debt Redemptions ($110.529B) – Tax Refunds ($0.131B) = $9.169B od Cost

Then, we do the math…$3.449B – 9.169B = $-5.720B

So our daily deficit is $5.72B. Yikes!! How much money did you lose 11/8?


First, some background. I’ve been tracking the US federal debt for about a decade now. I started off using the Treasury’s “debt to the penny” site, checking in every month just to see what the monthly change was. Although I knew there would be timing differences, I just assumed that delta debt was a fairly good indicator for monthly deficit. Unfortunately, for a lot of reasons, nothing ever tied to the published deficit numbers, and during the 2008 fiscal crisis, the correlation just went nuts. So, I dug into the details, and stumbled across what is known as the Daily Treasury Statement , or DTS.

The DTS is a little nugget of gold…essentially a daily cash flow statement of the US government. It has hundreds of daily data points…cash in, cash out, daily, monthly, year to date. It is the financial history book of the United States…and they go back to 1999, so we can compare a few of those “surplus” years to today. To me, it is all very interesting. I could probably write a book about it, but let’s be honest…nobody reads books anymore, so i have decided to live blog the debt. Perhaps not as interesting as an election, but as an accountant, the number of semi-interesting things I have the expertise to blog about is quite limited….