All of that is a nonsense and a distraction. The tax system is completely arbitrary and made up. It doesn’t matter whether something fits a certain definition or not. The only thing that matters are what the benefits and what the costs of doing a certain thing is.

And it’s pretty obvious that there are almost no benefits and there are massive costs to this approach.

could you disambiguate to which approach are you referring when you write "this approach" ?

https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...

tl;dr - previously companies could choose 1 or 5 years to amortize, now no choice & clarification that most software development counts as capital improvements (like building a factory). Many smaller companies are getting surprise tax bills this year, many of us have been putting it off with standard tax deferments (6 months)

https://blog.pragmaticengineer.com/section-174/ | https://news.ycombinator.com/item?id=38870429 (blog post by same Twitter user, Gergely Orosz)

https://news.ycombinator.com/item?id=38642461 (older tweet from Gergely Orosz)

https://news.ycombinator.com/item?id=34627712 (largest Ask HN discussion)

Also like building or buying a set of tools

People here love to complain about this rule but they generally don’t have any reason to justify themselves other than “it’s bad for me personally”.

You’re building software. That software is generally intended for use in the future. It is not “consumed” this year. Youre not deleting your repos every year. You’re a tech a company. Your product is an asset that you built.

There’s an argument that maintenance could be carved out but generally speaking most development is development. Triply so for startups.

The startups crying that they shouldn’t have to pay taxes until they’re bigger can deal. The only reason you have a tax bill is because you have taxable revenue.

It gets interesting when you think of responding to an outage. Are you "repairing" something that is broken, or "fixing" a bug and improving the software?

It's hard to analogize what we do to the historical context in which the laws were written.

> The startups crying that they shouldn’t have to pay taxes until they’re bigger can deal.

The same amount of taxes get paid, it's just about when / how it is spread out. The rule change means that the youngest and smallest companies have to bear a significant burden. For these companies, the lifespan of software is much shorter, and is arguably much more "researchy" in that we are trying out lots of things to see what works.

Large tech companies like MSFT, GOOG, APPL, ... they can manage this because they have the expert accountants and cashflows to weather the change over

Maybe you aren’t. Corporate America can be extremely inefficient about building software.

Actually the justification is pretty simple.

Tech startups are the future of the country and the world. That’s how much of the wealth has been built in this country for the past few decades.

And there’s no other country in the world that taxes this way, which means that every other country has an edge over the U.S. now when it comes to tech startups (or frankly, any startup where the majority of the cost is human capital).

It’s even worse in Tech because it’s so easy to work remotely so there’s no loss to establishing a company in the UK for example.

Finally, if a country does impose a certain taxation disadvantage to itself, it’s usually a trade off against some benefit. But it’s also hard to see what the benefit is here because the government is still getting the same money over a few years span.

So it’s dumb because the impact is entirely negative. That’s why no other country has done such a dumb thing either.

That’s certainly how founders and VCs like to justify siphoning so much money into their pockets but it’s important not to confuse share price with value. Most of the big hits were companies like Uber or DoorDash whose greatest success was reinventing old ideas with enough backing to evade law enforcement, and while they’ve made their management wealthy the positive impact on society is pretty mixed: marginal but expensive improvements in convenience with worse conditions for workers and large downsides for everyone else (e.g. Uber made traffic and pollution significantly worse and cut into transit ridership), and they’re the one in a hundred successes.

I don’t think startups are terrible but let’s not lose track of just how much money was poured into concepts which were clearly incapable of turning into a successful product (e.g. the billions spent pushing blockchains for things they’re bad at) or, especially, the ideas which might have gone somewhere if people were running lean instead of trying to cash out in the VC lottery.

All it does is frontloads the money. So if your company lasts beyond 3 years you’re in basically the same position. But those first 3 years you’re far worse off. And the government makes no additional money either because the same amount is deducted, just over 3 years instead of 1.

Second, any startup that shuts down will never pay the taxes on the revenue they earned but nullified with expensed product building costs.

Depends where, really.

In Austin (for example - but this applies to a lot of the US I suspect), it had no impact on transit because there was no worthwhile transit to have an impact upon. Instead it made drunk-driving rates plummet.

Should it be the small, innovative companies that made & make America great? Or the government who spends too much money as it is?

One of the government's jobs is to set rules such that they incentivize certain behaviors. The tax rules can be modified to a middle ground. Put some threshold in $X millions of dollars where you no longer get the choice and let the innovators leverage the money to build more great things. Big tech has the cashflow to weather the transition from 1y to 5y. Companies who spend every dollar that comes in are having a hard time. It means hiring fewer people in one of the most important industries

It's only large capital expenditures (that usually can be resold for similar amounts as purchased at) that require amortization/depreciation. For example, an Excavator.

Same if you build a house, building, etc. you can expense the labor that went into it that year, as well as materials consumed in building in. [https://onekeyresources.milwaukeetool.com/en/when-to-expense...]

There are situations the labor would be considered a capital expenditure, which you can then get ROI off of, but amortizing labor costs over multiple years is a pretty odd situation.

If you couldn't, it would get reaaaaly ugly real fast, since the big challenge here is that cash gets spent this year, while it is only deductable against income taxes years later. So if I get $1mln in income now, and spend $1mln on say software R&D, i'd end up owing a bunch of taxes on the $1mln of income using cash I just don't have anymore.

In the case of an excavator or whatever, there is a sellable asset which someone can turn around and sell to get cash if there is a problem. In the case of construction labor, there is usually some improved property or accounts receivable somewhere that can be used to pay the taxes too. In software R&D there rarely is.

Outsourcing can do it though.

Then go get a loan which should be easy because you’ve demonstrated $1M revenue with 100% gross profit margin. A complete non issue and painfully unrealistic because how are you paying for the $1M capitalized expenses before you have the revenues? Yes, you have to raise funds to start a business. Shocking.

On a cash accounting basis it’s actually worse, because that company is seriously in the red after paying taxes with the new rules. Previously it would have netted out at $0.

Of course someone is going to be hesitant to loan money in that situation, unless you can cut those expenses. Which if you’re in a highly competitive market could mean death to your company.

Is it overly simplistic? No shit, it’s an example to illustrate why startups are getting worked up right now because they have sudden cashflow issues. And why larger companies are also doing layoffs and trying very hard to reduce their software dev spend. Including outsourcing, where they can.

Because the actual way most companies will solve this once they figure it out is to not spend all their cash on software dev, and instead reduce it significantly.

Because this new treatment means instead of being able to write income off as a straight business expense in the year they spent it, they have to pay taxes on money they’ve already spent in that year, and then can only make it up fractionally after that.

So in the example I gave? Companies taxes that year are zero with the prior R&D rules. They had zero taxable income.

With the updates to section 174, now they owe taxes on $800k of taxable income the first year.

Huge difference.

They can still eventually write off the whole $1mln they spent in year one, but that is a hell of a big difference in cash flow. Even if dev spend is flat for 5 years, it will take that long for it to look the same cash flow wise. And it’s rarely ever actually flat.

And that is not even taking into account the increased cost of money going on, tightening access to credit. If you pulled off this situation while being in a safe industry that a bank will write a loan for you easily, then cool.

But they’re surprisingly weird sometimes, and if you’re in year one it can be hard to get anyone to return your calls at all if there is anything ‘non-traditional’ about you. Which is most startups.And the ‘startup banks’ where it was easy (SVB and first republic) all got nuked.

So you could easily end up paying 25% APR on short term financing to close this gap, if anyone is talking to you at all.

No, it’s not. Net income here is $800k (20% amortization). Gross income is $1M. Because accounting recognizes that the value of your investments today are useful next year. If you sold a million dollars of your software this year, it’s extremely likely that you will be able to sell the same software next year. You’re not starting from scratch in 2025.

> On a cash accounting basis it’s actually worse, because that company is seriously in the red after paying taxes with the new rules. Previously it would have netted out at $0.

And that’s an example of why cash based accounting is not they way businesses track their financials. It’s misleading.

Your understanding of finance here is completely backwards. Think about it. Would you rather loan money to a company that paid $1M contractors to do work for $1M in revenue? Or to a company that spent $1M building a product that sold $1M in services that year?

Obviously, OBVIOUSLY, the second one is more enticing. The contractor company is extracting zero profit and is poised to do the same next year. The software company built a product and they can continue selling. Whether the market is really competitive and whether they might need to continue investing in the product to stay competitive is a tangential concern and a secondary consideration about the business environment that does not factor into our basic understanding of assets here.

Whining about funding being hard to get or that business is hard does not define what an asset is. Imagine a factory owner saying “well the Chinese just built an automated factory so I’ll need to double my investment next year to make sure my products can compete in the market, so I’m going to say that my building of factory doesn’t count as an investment because that just works out nicer for me. It’s only fair. The business environment is competitive!”

Because in theory sure.

Though previously there was the far more profitable and easy to deal with theory of building an actual asset of large value (monetized via stock/equity) while treating the expenses to do it as straightforward business expenses with no amortization to be a PITA, and the IRS was willing to go along on that.

In practice, this is exactly how you run your startup into a wall at 100 mph while you bounce payroll checks and end up in a morass of legal trouble. Because cashflow absolutely matters, and startups and small businesses rarely can just get ‘free’ debt at a moments notice to smooth it all out. Even large ones are having more difficulty than ever.

Call it whining all you want - it’s a significant concern partially driving market behavior. Remote work also meaning outsourcing is back on the table being another.

“Business is hard” is not an argument to redefine what an asset is

Which most banks or investors are going to be pretty hesitant to do, because cash crunch with low assets?

Some business environments, sure. Others? Not so much.

Is it impossible? Hopefully not. But is is a major change from how it used to work.

Which is why there is such a huge push for offshoring.

As stated, without some significant changes that company is doomed in short order with the new rules.

Not ones that you build.

For it to be a capital expense they'd need to be large, sellable assets. For example, formwork? Jigs (except for something really exceptionally expensive perhaps)? Hand tools? Less-expensive-than-a-truck custom machinery? Not a problem to write off as a normal expense.

'Gigapress'? Yeah, thats a capital investment/expense.

The difference here is 'is this something you built/bought to get a job done' (aka an expense), or 'is this a large valuable asset (with a market value)'. It's a sliding scale with many exceptions.

Someones $100k snap-on set for example MIGHT be considered a Capital expense, not a normal business expense, depending on the circumstances and scale.

There is a pretty sizable cap too where you can do whatever and not worry about it too much, a bit over $1mln if I remember correctly.

What changed is that almost all software development was classified as capital improvements and a forced 5y (15y if international salary). There is not way to classify different software projects by their useful lifespan.

Section 174 was updated this way yes - which is why we’re talking about it yes?

Previously it was classifiable as an R&D expense, and could be written off the same year it was taken. Which is much friendlier, cashflow wise.

Anyone manually depreciating computers over 3 years either already has a lot of capital expenses (> 1 million/yr), or should probably get a better accountant.

There are situations it needs to be more complicated - but for the vast majority of the situations anyone reading this is going to be in, it isn’t. And if they are in that situation, they should hire a professional to help them untangle it.

The salient thing here is how long they are expected to last. Super cheap items are generally not going to last very long so yes they do trend towards expensing.

There is ABSOLUTELY no requirement that a capital asset be sellable or have any market value.

You get an F-

[https://www.law.cornell.edu/uscode/text/26/179]

“ (a) Treatment as expenses A taxpayer may elect to treat the cost of any section 179 property as an expense which is not chargeable to capital account. Any cost so treated shall be allowed as a deduction for the taxable year in which the section 179 property is placed in service.

(b) Limitations (1) Dollar limitation The aggregate cost which may be taken into account under subsection (a) for any taxable year shall not exceed $1,000,000.”

[https://www.executech.com/insights/1-million-deduction-for-n...]

For something to be considered an asset, it has to generally be “property owned by a person or company, regarded as having value and available to meet debts, commitments, or legacies.”

While there are certainly edge cases where something would be impossible to sell or value on a market but still considered an asset; that gets pretty murky for a number of reasons no?

An asset which no one would ever buy is also going to be hard to consider something which someone would take as payment for a debt, meet a commitment, or leave as a legacy. Even a pile of beany babies has some articulable value.

If it isn’t an asset, it is likely an expense.

Most businesses do not account for or attempt to track the things I’ve been calling out (hand tools, etc.) at the level they consider them assets. Any more than they track a desk or whatever. It’s de minimus.

Would they get worked up if someone walked off with them? Sure. But they aren’t depreciating their stocks of beaten up wrenches or counting their hammers each year, anymore than they are doing that for desks.

Unless they have a really bored group of bean counters or are so large they have no choice, because they literally bought like 10k desks and now that is real money.

Excavators, bobcats, etc? Yeah. Everyone is tracking those unless they’re really screwing up their taxes.

Assets do not need to be sellable. Goodwill is an asset. It’s a not “real” thing. The book value is also completely unrelated to the market value.

You have quoted a dictionary definition of the word “asset” from Oxford languages off of Google. You do realize we are not talking about about generic language but specific technical accounting definitions right?

Desks are 100% capital assets. Office furniture has a specific useful life defined by the IRS. Feel free to Google this basic information.

That exemption means you can either track them as capital assets and depreciate them, or take them as straight up business expenses in the year you qualify.

It’s right in that first paragraph I quoted. Bullet a.

Aka ‘whatever’. You would need to decide though.

If you have a different definition of asset that includes ‘something of no market value’, then feel free to provide a reference.

If you buy something you can’t sell later for money, it isn’t an asset anymore. Its economic value has been consumed, and it was an expense.

That is not true in the U.S. All capital assets must be depreciated. This includes all tools and equipment except for very short life consumables (< 1 year).

Why do you think they should? Care to quote a section of the tax code?

No one, including the IRS, wants to deal with mountains of BS paperwork over small stuff like this.

For a description [https://www.trustbgw.com/blog/2016/07/08/capitalize-fixed-as....],

“To determine when to expense and when to capitalize, consider both the de minimus rule and the useful economic life.

The de minimis rule allows you to expense any item that may potentially be capitalized so long as said expense (or the sum of related expenses) does not significantly distort your bottom line. In other words, these expenses cannot make up a large percentage of your total expenses, subsequently providing you with an extraordinarily low income. A strong metric to follow is that the sum of the expenses should be less than 0.1% of your gross receipts for the year, and/or 2% of your total depreciation and amortization expense for the year.

Any item that will retain no value after 12 months may be expensed. Do note that the IRS may frown upon deductions that appear to take advantage of end-of-year timing. If you incur an expense in December and then consume the item in January, you may need to argue that the deduction did not distort your bottom line in order to take it.“

The de minimus rule is, as your own quote indicates, for a tiny, tiny subset of your assets.

That is paperwork. Unless your tax folks are just yolo’ing amounts into the appropriate fields in your taxes anyway, in which case you’ve got bigger problems.

You are correct that it doesn’t necessarily require going out and tracking say every individual desk.

But to depreciate it, you would need to be able to show you bought x desks for y $$ (and those were reasonable business expenses), including receipts, and if the IRS got worked up, they might demand to see said desks. Which if you had no idea where they were, is going to be hard eh?

And then you need to track them over the period of time you’re depreciating them (and how much was depreciated).

Which for a small business is not worth the hassle at all.

And it is highly unlikely the amount they spent on said desks is not de minimus.

If you are IBM or sprint or whatever? Then sure. That is why they have massive accounting departments, asset tag everything, etc.

Even they don’t bother with hand tools and the like for tax purposes though.

they can’t do things the easy (and legal) way.

Just seeing if I got this right. A McDonalds' flipper's and a SWE's wage expense now works the same tax wise?

I mean, most of the R in RnD is abuse for tax reasons when applied to SWEs. I get that it is bad for our wages, but is the change 'fair' for some definition of fair?

Exactly the opposite

A burger flipper is the cost of doing business. The value of the burger flipper is used immediately and so the expense is recognized in that year.

Software used to be treated like that. Now they say that software development is more like building tools and developing assets. So you should recognize that expense’s value across a handful of years.

It gets interesting when you think of responding to an outage. Are you "repairing" something that is broken, or "fixing" a bug and improving the software?

It's hard to analogize what we do to the historical context in which the laws were written.

Not sure who benefits from this.