$3 Trillion in Forgotten Debt | Hacker News Reader

$3 Trillion in Forgotten Debt

bloomberg.com

229 pointsgscott63 comments

If you're accounting for future lease obligations as debt, then I think you somehow have to account for future exclusive use of the thing you are leasing as an asset.

It's easy to value the future lease obligation -- monthly payment * remaining term. But without a way to value the corresponding leasehold this change seems like it just distorts financials in a way that makes the data less useful.

It bothers me that I have Ph.D. yet I understand only about every third word in your comment and nothing of the whole thing.

I see what you did there ;)

He said if you are going to count the monthly rent against a renter, you have to count the value of the place to live too.

If I (a company) sign up for a 10 year lease on a retail space, at $10,000 per year, it's easy to see that it's a $100,000 obligation (ignoring things like present value of money etc).

Apparently (according to the article; I'm not an expert on this) that $100,000 obligation does not appear on the balance sheet. The new IFRS 16 rule means that it will; so companies will now have a lot more debt. Grandparent commenter was asking "how do you assign a dollar value to the exclusive-use value of that lease" - for example maybe the neighborhood where your shop is suddenly became really popular, so you're locked into a below-market rent for the next 8 years.

It's an interesting objection but I don't see why you'd have to assign a dollar value to it. Yes you might be overall doing well compared to other possible options but that doesn't change that you're still owing $100k.

Perhaps I'm not understanding the objection but it seems like an unfounded one.

You owe $100k, but you also have something of value (worth approximately $100k when the lease was entered, but should be adjusted for transaction costs of subleasing). Over time, your payments reduce the debt (this balances out), but the asset also reduces value, which reduces your overall balance. Adjustments for marketability of the leased property should also be made.

If the property isn't marketable, maybe the current asset value is $0, but that's rarely the case.

Hmmm... it seems to me to be identical to how home owners have a mortgage and home value as two separate debt vs value measures.

But what your describing doesn't happen in they way you're describing for home owners why should it for someone leasing a property?

No accountant, but here is an explanation:

Say I open a sauna in a desert. Probably not going to pay any rent, probably not going to make any money. Under zaroth's suggested way of looking at the situation this something like $0 debt, $0 asset.

Say I open a sauna in a luxurious ski resort. Now maybe I have some huge 'debt' of rent * lease lifetime; but I also have a productive asset worth some hopefully large (the right to put the sauna somewhere it will get customers).

If we only account for the debt only, we are at some level asserting that opening the sauna in the desert was a better bet because the rent is lower. This is clearly stupid, the rent is half of a transaction where we are claiming value.

If the business owner puts the upcoming lease payments on their books as a liability, then to keep the picture clear they should also be adding a productive asset to the books that represents their valuable location. If they are operating a rational business, it is likely that the asset added is equal to or greater than the liability.

The productive asset IS accounted for in the income/revenue... surely?

If it is not producing revue then it is clearly not productive.

Well, OK, so this is one of those things that confuses me as a non-accountant. The problem is that you're conflating two things: Debt (as in the amount of money you have to repay) vs something like "value". Why call them the same thing?

> If we only account for the debt only, we are at some level asserting that opening the sauna in the desert was a better bet because the rent is lower

No no no. This is when you misread "debt" as "value". We're just saying that the desert situation is less debt. The ski resort situation is more debt. You can be in more debt but overall have a more profitable enterprise, right?

> If the business owner puts the upcoming lease payments on their books as a liability, then to keep the picture clear they should also be adding a productive asset to the books that represents their valuable location.

This again confuses me - the debt is a sure thing. You signed a contract, meaning you have to pay up no matter what. The productive asset is not a sure thing - there's a million ways you could mess up and make less money than you intended. So it is absolutely fair to say that the desert business is low-risk (you're going to lose $0 in the worst case), and the ski resort business is high-risk (you're going to lose a lot more in the worst case).

If guessing the future value of productive assets was simple, then investors of all kinds would have a super easy job. So how could we possibly put the productive asset in the books with a straight face? Isn't that borderline fraud? I can make my projections as optimistic or pessimistic as is advantageous to me. Who's to say?

replace the ski resort with a money printer. That money printer has value, given it can print money.

You can also assign a dollar amount to how valuable that money printer is. So you can justifiably stick it on your balance sheet since this is a way you're making money.

There's a whole thing in accounting about assigning dollar values to things you own, here it's no different. I would argue that you shouldn't really assign potential revenue from the ski resort, but depending on your circumstances it _could_ make sense.

Accounting is a bit arbitrary in the end, since there are multiple logical ways of representing things. In most cases it's impossible to claim that one way of thinking is wrong... at worst you can claim that it doesn't match accepted practice.

This isn't mixing debt and value. The core point is that if the business is taking on a debt we should have regard for what the debt is being spent on, and the principle is that taking on a debt shouldn't immediately reduce equity in a business because it should be balanced by an asset.

> You can be in more debt but overall have a more profitable enterprise, right?

Yes. But if the businesses are identical in every way (management, worker skill, etc, etc) then they should have identical earning potential. If they are different in a way that is profitable, that difference is an accounting asset.

It doesn't matter how similar the hypothetical sauna in the desert is to the profitable sauna, the better location will result in better profits.

> This again confuses me - the debt is a sure thing. You signed a contract, meaning you have to pay up no matter what.

No. Nothing about the future is certain. I don't even mean that in the pedantic philosophical way, it is a very practical concern. If, eg, my fictional ski resort was buried in an avalanche that the rent would be extinguished (in sane jurisdictions, anyhow). The business might also be limited liability and go bankrupt - same $0 loss worst case as the desert.

> there's a million ways you could mess up and make less money than you intended.

That is true of any asset. Risk is an accepted concept. Even 'risk free' assets are only risk free assuming the law doesn't change.

> If guessing the future value of productive assets was simple, then investors of all kinds would have a super easy job. So how could we possibly put the productive asset in the books with a straight face?

Because 'we' think it has value. It passes the 'someone with money is paying for it' test. Guessing values isn't easy, but at some point we do have to do it to compare options.

> Isn't that borderline fraud?

I'm not going to try and lawyer and account at the same time, but accounts books aren't some magical tome that is always right. Sometimes erroneous valuations make it into them. The accountants do the best they can.

There are no set of accounts so correct they can make things true. There are reasonable and unreasonable assumptions. "I think this lease is net-positive for equity" can be a reasonable assumption. In fact, it usually will be.

If I buy a computer for $2000 it goes on my asset sheet at $2000 and depreciates over time according to GAAP rules. Shouldn't my lease contract on a $1000/month 5 year contract be a $60,000 asset and a $60,000 liability with the asset depreciating at $1,000/month? In most of accounting we get double entry bookkeeping to work by assuming the cost paid for an asset is the value of the asset. I think in many cases reporting it this way would be more accurate (Although depreciation expenses would be a giant loophole that needs fixing).

Hi there i just want to add that it is the present value of future payments that are put on the books as 'debt' (i.e. liability). Also, In situations where you truly do have a 'cheap' asset relative to the market, you are able to value that separately and many companies do. Search "below market leases 10K" for further information.

Lease accounting guidance is very complex and there are many nuances that cover concerns in this thread. The point of the guidance is to determine the fair value.

My take on the comment: Compare to a family that takes out a loan to buy a house. If you ask their net-worth it's not just money in the bank, you have to subtract the debt that they owe. Parent comment was saying that you need to add the value of the home back in as well. It's a little different with leased property, but there may be some validity to the comparison. My analogy is not the same - a family with a mortgage will eventually get rid of that monthly payment and own the home. A family renting is not going to achieve the same thing. In the same sense I disagree with the comment - the property leased is not an asset, but it does have some value or it wouldn't be leased in the first place.

OTOH the article mentions that for a while companies liked to sell and lease-back which allowed them to appear to have cash even though the cash didn't have anything to do with business activities.

A lease or rental agreement can be valuable depending on the terms.

An example: my mother managed to significantly negotiate down her commercial lease after a court case with her landlord.

The price that she now pays is significantly below market rate. It’s also locked in for the next decade or so. That lease is quite valuable.

The same is true for rent controlled apartments. I have had one in San Francisco for the last 5 years and pay significantly below market rate.

You probably shouldn't be unless your PhD is in Accounting or Finance.

What word in the comment is difficult to understand?

I am also confused as to what having a Ph.D. has to do with your ability to understand concepts in what I am assuming is a completely unrelated field?

A Ph.D. signifies advanced knowledge in a particular subject, not a generalized superior ability to understand unrelated concepts.

It’s absolutely astounding how much people think a phd means.

Fwiw all you need is an ability to google. You can find google at https://google.com.

That seems to be what it's already requiring:

"Under IFRS 16, a lessee is required to recognize an asset for the right to use the leased item and a liability for the present value of its future lease payments." https://en.m.wikipedia.org/wiki/IFRS_16

At meta level, it would be interesting to know why IFRS chose the philosophical position that the lease remaining unpaid term is a "debt".

It seems like another valid way to look at the lease (for renting space) is to consider it a "call option" on a agreed upon price with an expiration date. Therefore, you don't count the whole lease as a bank-like debt but instead, count the "lease termination" fee.

(Or maybe I'm taking the examples discussed in HN too literally and the actual IFRS text already covers both situations and the company can choose to record "remaining term" or "termination fee" -- whichever is the lower amount. The termination fees for equipment like airplanes and tractors is higher percentage (must pay remaining residual value) than office space which makes it behave more like a "loan".)

if there's a way out of paying it other than default that's an interesting notion. option-pricing models are a bit more complex than accounting seems to like though. seems like an opportunity for a data-vendor or hedge fund: option pricing of leases for improved valuation models. hmm...

For accounting you probably want to keep it simple. Accounting should be a baseline sanity check to keep numbers somewhat in line with reality. It's not meant as a valuation model.

>It's not meant as a valuation model.

That's not my intention. I'm staying within the bounds of plain accounting for the concept of "debt liability".

If a company has a 5 year lease of $1000/month rent, I think the new accounting rules says it's a 60x$1000=$60000 "debt".

But in my mind, the company really isn't on the hook for all $60k. If the company has an "early termination" or "lease buyout" clause that they can exercise at year 2 with a 6-month notice to the landlord, they are really only in "debt" for 30 months (24+6), not 60. The rest of the 30 months act more like a "call option" than a debt instrument because the company really doesn't have to pay it. I'm not an accountant and just thinking out loud. And to reiterate, this "optionality" math applies more to office space leases than equipment leases like airplanes.

In general, there are a lot of contract terms that affect your flexibility as a business, and they are often major negotiating points for large deals. Almost any of them could be considered a call/put option.

But I don't think the goal of accounting is to expose all of those contract terms to casual investors.

If you ha these terms the debt would be for 30 months. Quite clear from IFRS 16, only the amount which you are contractually obligated to pay is debt.

> is to consider it a "call option"

A lease is really closer to a set of futures contracts than a call option. You can't just walk away from a lease if you don't want it anymore.

The company walks away by paying the early termination fee. We assume that type of clause is often negotiated into the contract because the company wants flexibility to break the lease to respond to changing business conditions.

I don't know about corporate leases, but early termination fees on a lease AREN'T a given. Some states have next to no regulation on it, and landlords are free to write terms without one. The early termination fee there is simply "the remaining rent", which is what is reflected here.

>but early termination fees on a lease AREN'T a given.

Correct, and that's why I specifically wrote "often negotiated" and not "always included". I wasn't writing in absolutes.

>The early termination fee there is simply "the remaining rent", which is what is reflected _here_.

(Not sure what previous comment the "here" in "reflected here" is referring to.)

I was talking about leases where the negotiated termination fee is much less than the remaining rent. It depends on market forces. (If the office landlord has lots of vacancies and is desperate for tenants, favorable termination fees will be more common.)

If not putting leases on balance sheets understates actual liabilities, then likewise, adding entire lease terms with optionality overstates it.

Since you have to balance asset/liability entries, if you try to model out the cost of termination (which can include TI allowance refunds, commissions, fees and other not so straightforward cals), you might end up with an unrealistic asset picture where the lease is showing lower overall cost than it should.

It is required because bad actors historically manipulated lease accounting to misrepresent their contracts and future payments they are liable for. It is not 'debt' - it is a contractual liability, valued at a fair and reasonable value (including valuation of termination provisions).

> is to consider it a "call option“

Early termination is an example of optionality. Callable bonds and demand deposits also contain optionality. That doesn’t make them options.

>That doesn’t make them options.

Yes, I know. I was using <quote>"call option"<unquote> as analogy and not a strict legal financial instrument. Sort of like the "Greenspan Put" isn't really a "put option". Or how some might saying investing in Uber is a sort of like a <quote>"call option"<unquote> on future taxi & employee regulations favoring companies like Uber.

I would like to caveat by explanation by saying I am admittedly less versed in IFRS than US GAAP, so there may be some technical differences between the two, but the spirit should be the same - there is actually corresponding gross up of your assets, equal to the net present value of your minimum lease payments, when you put the lease obligation on your balance sheet.

So at the the time you sign a new lease, you will put an asset and liability of equivalent value on the books. Assuming the useful life of the asset extends past the lease term, general expectation is that the asset value would be amortized over the term of the lease, and at the same time, interest expense will be recorded on the lease obligation. Ultimately the same amount of expense will be recorded (equal to the minimum lease payments), but the timing of the expense is different as it front loads the expense (because of the interest accretion). So you have that and the balance sheet gross up that will be different. The change, in theory, provides greater transparency related to what can be a significant obligation for a company.

> the timing of the expense is different as it front loads the expense (because of the interest accretion)

I would expect that if a firm could make a good case to its auditors that the revenue derived from the use of the leased asset was expected to be higher during the early years or lower during the later years, or vice versa, then the auditors would allow a timing adjustment somewhere.

If my corporation has just signed a 12-year operating lease for some property in Los Angeles with the intention of profitably obtaining revenue from the property only during the 2028 Olympics, do these accounting rules lock me into such unrealistic reporting?

Many years back, I spent quite a bit of time on similar but different issues of recognition of revenue and expense, and it seemed that if one could make a good case, there was flexibility, but everyone wanted to abuse it. Is the flexibility still available in reasonable cases, gone now in general, or gone only for specific classes of contracts like operating leases?

There is such a thing as a "capital lease", or at least there used to be. Capital leases work in this way--both asset and liability are recorded. It's only fair that the new rules for operating leases works this way too.

So does this mean we can depreciate things like 3-year AWS reserved instance costs, as they are now counted as assets? Seems pretty nice to me.

Switching from expensing a reserved instance cost to capitalizing and then depreciating it is actually a bad thing. It increases your short term paper profits and decreases your long term paper profits, but doesn't change your pre-tax cash flow. Since you pay taxes on paper profits, it increases short term taxes and decreases long-term taxes, reducing after-tax cash flow in the short term and increasing it in the long term. Since money has time value this reduces the value of your business.

In general, if you have the option for tax purposes, it's better to expense something than capitalize and then depreciate it.

Remember that leases aren’t just leases for real estate. Many companies operationize most or all capital costs. Companies dump all sorts of crap into them.

When I used to sell stuff in college, my fortune 50 customers dumped all sorts of stuff into leases like office supplies, etc. My account had me running around buying thousands of pens from stores around us to for one Fortune 50 customer to roll into a lease for a bunch of servers.

When does this cross the line and become fraud?

Who would be the victim of fraud in the situation described above? It's my understanding that both the lessor and the lessee know about the office supplies.

And the IFRS rules in question came about to ensure investors were similarly informed. Leasing is less capital intensive, more tax efficient and more flexible than owning.

Case in point: when I consume a bottle of water from my Uber driver, I am consuming it as part of a lease.

The investor.

Usually leases allow for overhead costs to be lumped in to some percentage.

If you buy something, you can roll installation services into it. This was just stretching it.

Unfortunately many leased items are rarely able to produce an income or realise a "sale price" if the original purpose is gone.

Thus more often than not, its asset value is almost impossible to reliably be assertained by the business let alone by an external party.

For example if all these aeroplanes i have have an income producing capability of producing n $millions but my business can't use them at a profit because of competition, how are the aeroplanes "asset" value ever going to be realised? they are effectively worthless.

So the balance sheet should show them at a high figure when business is good, and when business goes down the toilet they should be written off as they can't even be sold.

This makes less than zero sense

David Stockton's been pounding this for a long time, he pointed out that Toys R'Us was the first canary to fall. His point is, cheap credit has allowed many unhealthy businesses to keep sputtering along for a long time without making changes.

Granted, Stockton is kind of a gold bug, and former Reagan budget director trickle down fan, but he's reformed on the debt issue and taxes now, and has been basically predicting an apocalypse once central banks start cleaning up their balance sheets and tightening interest rates.

> he pointed out that Toys R'Us was the first canary to fall. His point is, cheap credit has allowed many unhealthy businesses to keep sputtering along for a long time without making changes

That's an odd point to support with Toys R Us? From what I read, a third party borrowed money to buy the (profitable) Toys R Us. The new, indebted, Toys R Us is still profitable except for the cost of servicing the debt that, notably, it didn't choose to take on. Therefore it's going bankrupt and the buyers are getting wiped out, but the store will be fine because its business is perfectly healthy - revenues exceed cost of operation.

This promises to be a nightmare for small business. By way of a few examples, a company will need to evaluate whether its right to use tangible personal property is actually its own property subject to personal property tax, sales tax and use tax. Generation transfer and businesses sale valuations will be impacted on a perishable lease "asset" values. There will be disputes about the differences between Fair Market Value and GAAP value. Even colocation, SAAS and software leases with terms greater than one year may be impacted by this rule.

How much will this shift the Capex verse Opex dynamic which typically is perceived to be one of the many motivating factors towards cloud computing adoption?

The cost to own and manage is still greater, but it for terms longer than one year it will likely raise costs to lessees, and will certainly impact their balance sheets.

This will make debt more attractive for companies that produce something. At the moment the fact that leases are not visible drives productive companies to rent equipment instead of renting money. The big beneficiary were GE Capital and the like that in turn rented the money often at a discount. Leasing companies will continue to enjoy some economies of scale but have lost a strategic advantage. Any effect this rule change will have on companies will show up magnified in the leasing companies.

https://en.wikipedia.org/wiki/Hollywood_Accounting FTW

As an example of how accounting rules can affect investor perception:

When Apple started selling the iPhone, they recognized the revenue for each phone over 2 years. They thought they had to do that in order to provide free software updates.

Within a couple years the accounting rule changed and Apple could report full revenue from an iPhone sale upfront. Suddenly their GAAP earnings jumped, which led to a stock jump too.

... They haven't made the change yet:

https://www.marketwatch.com/story/apple-changes-tune-on-new-...

> “The Company plans to adopt the new revenue standards in its first quarter of 2019 utilizing the full retrospective adoption method,”

It won't have much of an impact as the deferred revenue is accounted for and is not the full price of the phone, but the value of future software updates ($25):

https://www.sec.gov/Archives/edgar/data/320193/0001193125100...

> For all periods presented, the Company’s estimated selling price for the software upgrade right included with each iPhone and Apple TV sold is $25 and $10, respectively.

In layman’s terms, does this mean that some company a) borrowing money for 10 years to buy machinery, and b) leasing that machinery for the next 10 years, will be reported equally on its balance sheet?

More or less. Big difference is that 10 years leasing would still be less debt/asset than owning (cars being a prime exampel with leasing usually being for 3 years at a time, obviously a smaller asset/debt than buying the same car).

By the same reasoning, wouldn't employee salaries be a contractual liability? Though without committed end dates, usually.

No, because as you said there is no commitment. Pension plans, however, generate liabilities if underfunded.

Debt is a tool. It can be forgotten or demanded. It's the leverage a few exert upon the world.

Not all debt is created equal.

If you are opening an airline, you need to buy aircraft (amongst many other things). (Commercial) aircraft are extremely expensive but generate generate value long after they are acquired. It is a lot cheaper to incur devt by leasing those planes than it is to buy them outright.

Personal debt is different, in most cases, and is a lot closer to the reality of your statement. I’m spending about $11,000 for our wedding, all inclusive. I paid for that with a credit card. I can’t make revenue off of that wedding outright (though I guess it can become an asset if an epic YouTube video comes out of it and becomes a viral sensation that survives the test of time), so that debt isn’t useful. But I made a risk calculation that me paying for everything upfront and paying for it later is more important than waiting the n years it would take to save for that wedding and pay everything in cash. That decision _could_ bite me in the ass in a few years when IT is 100% automated by ML/AI putting the cloud into the blockchain, but I made that decision assuming that that wouldn’t happen. :-)

Reply on news.ycombinator.com