Hacker News new | past | comments | ask | show | jobs | submit login
How to Detect Business Fraud (economist.com)
169 points by known 5 months ago | hide | past | favorite | 68 comments



For those who are interested in the accounting side, I highly recommend Financial Shenanigans by Howard Schilit [1]. If you don't have an accounting background, all you need to get value out of this book is a basic understanding of the double entry accounting method, and understand the four financial statements: balance sheets, income statements, cash flow statements, and statements of shareholders equity.

After reading this book, I found myself digging through earnings reports to look for signs of shenanigans, and have found cases that raise my armchair-accountant eyebrows.

[1] https://www.amazon.com/Financial-Shenanigans-Fourth-Accounti...


Thank you for this, I just bought the book. Any books you'd recommend to go over the prerequisites you mentioned? I read this [1] ages ago but I honestly don't remember much at this point.

[1] https://www.amazon.com/How-Read-Financial-Report-Wringing/dp...


I took a class specifically on this - how company manipulate their financial statements to make things look better.

It's really fascinating. If you're willing to dig into the financial statements (and ones from the past), you can learn a ton about how a company defines "performance" and whether or not it's reasonable.

Despite GAAP accounting rules, there is enough gray area for companies to hide a lot of bad information.


Any book you'd recommend to get a quick general grasp of accounting? Aside from the one referenced in the parent, which I'm probably going to buy.


The Accounting Game: Basic Accounting Fresh from the Lemonade Stand, by Darrell Mullis and Judith Orloff


Financial Shenanigans was mentioned by others. That one is pretty good!


Enron being the classic example of what you're describing.


Yup we covered Enron.

It's pretty interesting to study the mechanism of the fraud. Enron did a lot of things, but one of the big ones was how they booked revenue. Since they were a middleman, they should have booked their cut (i.e. fees) as revenue, but rather they booked the entire purchase as revenue, drastically inflating their growth.


If I remember correctly, at that time it was legal to book it this way. Perhaps that was "innovation" that allowed Enron to shoot to the moon?


So, what can you actually do with this knowledge?

Short their stock?

Raise awareness of their fraud, while holding a short position?


> Each stock fraud is fraudulent in its own way. But there are common elements. One is a breach between earnings as defined by Generally Accepted Accounting Principles (GAAP) and non-GAAP measures. Another is an increase in “days payable outstanding”, a yardstick of how long it takes a company to settle bills with suppliers. Delay boosts cashflow, at least for a while. So does gathering more quickly payments you are owed. Firms with dressed-up earnings also tend to pile on debt because they lack strong underlying cashflow. And there are grounds to suspect the worst of companies that engage in a lot of acquisitions. Aligning the accounts of acquirer and acquired gives ample scope for fiddling.

> Transcripts of conference calls with stock analysts can also be revealing. If the company keeps moving the goalposts, then be on alert.


Thanks, I couldn't read the article.



> 403 Forbidden -- Cloudflare



For others seeing this who need a short and simple fix: this seems to happen when your DNS server is 1.1.1.1, changing to 8.8.8.8 or anything else fixes it. archive.is has a disagreement with cloudflare at the moment which makes it unusable on cloudflare's DNS.


Add 94.16.117.236 archive.vn in your /etc/hosts


I don't think they actually use cloudflare. Neither their nameservers nor their server IPs are cloudflare. They only copied their interstitials.


> nor their server IPs are cloudflare

They are for me, apparently because I'm using cloudflare's DNS service, and cloudflare is returning its own IP when I query for archive.vn's.

  $ dig +short @8.8.8.8 archive.vn 
  5.196.68.232
  $ dig +short @1.1.1.1 archive.vn       
  1.0.0.1
  1.1.1.1
They don't return themselves for all hosts,

  $ dig +short @1.1.1.1 google.com
  172.217.9.14
I wonder by what criteria they intercept, and if this by archive.vn's or cloudflare's initiative.


This link should answer your questions and give the backstory.

https://news.ycombinator.com/item?id=19828317


The China Hustle[1] is a great documentary about the systemic securities fraud committed by Chinese companies selling into the American markets in a way that allows them to avoid much of the disclosure and due diligence typically required. It follows investors from the Muddy Waters firm mentioned in the this article and was made by the same group who did documentary version of The Smartest Guys in the Room.

I believe the film is currently streaming on Hulu.

[1] https://en.wikipedia.org/wiki/The_China_Hustle


Except, the problem, is that the China Hustle displayed AliBaba in a bad shape and I doubt the guys were short the NYSE:BABA stock. The film was slightly hinting that you should be shorting or offloading your BABA stock.

But what happened since? The film was released in 2017 (BABA around ~90) and now the stock is trading around ~195.

So if you offloaded your stock, you'll have lost on some nice gains. It's worse if you have shorted: It'd have been a real hustle and it's not clear when the stock will correct.


IMHO, exposing issues at the company does not guarantee its stock will go down. After all, investors (broadly defined, as in individual ones, 401k, mutual funds, pensions, etc) can keep momentum up. Something about the market can stay irrational longer than an individual investor's account positive.

So it is risky no matter what, just perhaps risk/reward ration increases maybe?


Thank you for this. The extend of these scams is mind boggling.


I think one of the most interesting methods of business fraud detection is Benford's law. It has been found that natural transactions do abide by this law.

https://en.wikipedia.org/wiki/Benford%27s_law


In real financial data you'll often see departures from the Benford's law. For example, if a company has $50 expense limit for dinner, you'll find a lot of invoices for $49.99. Same goes with approval limits at different management level. Not as useful in practice as in theory.


I've found the same. It sounds great and applicable but could not find any practical application with real data.

Even when applied to large number of real companies' journal entries, the amount of false positives is overwhelming.


Inflating sales is a classic move:

> In indictments involving the case, prosecutors said HBOC sold software or services to more than a dozen hospitals with conditional "side letters" that allowed the hospitals to back out of the deals. The side letters were then hidden from auditors and the transactions were reported as sales.[0]

McKesson (and perhaps other pharma wholesalers) also played games by timing invoice payments and chargebacks to increase quarterly growth.

0) https://www.nytimes.com/2005/01/13/business/mckesson-agrees-...

Edit: Another tactic is hiding rebate and chargeback transactions in numerous ~unauditable spreadsheets, stored on individual employee machines. Or even only as hard copy.


I don't know how such tactics as side letters are legal. In finance, there's numerous hoops and rules to jump through to actually prove that you sold and asset and did not merely engaged in a dressed up repo transaction.

a few links

https://en.wikipedia.org/wiki/Repo_105

https://www.deallawwire.com/2017/05/18/true-sales-a-refreshe...


Side letters are legal. Lying to an auditor is not.

Ideally, auditors would verify the legitimacy of a transaction with the counterparty. As in, physically visit their office and speak to the person whose signature is on the contract.

Unfortunately, this is impractical when there are hundreds or thousands of such contracts every year. There aren’t enough auditors and there’s not enough time. So mostly, auditors rely on the assumption that they’re not being provided with falsified documents, and they’re not being lied to.

Unfortunately, that’s sometimes (often?) just not true. Even though it’s illegal, it’s still highly vulnerable to exploitation.


They exposed Luckin Coffee $LK and called it a fraud in January to no reaction in the market. The stock collapsed 80+% before being suspended in April. They had this exchange of tweets with Citron Research which might be my favorite case of "I told you so":

https://twitter.com/muddywatersre/status/1245704324342108161


Highly recommend "lying for money" by Dan Davies, a very readable account of a wide range of business frauds.


Other great reads:

The Match King

The Smartest Guys in the Room

Bad Blood

Billion Dollar Whale

Also, famed short seller Jim Chanos teaches a class on frauds and recently posted this short list of recommendations: https://twitter.com/WallStCynic/status/1256962642499035137?s...


"The Smartest Guys in the Room" was great, indeed. Also, the book from the guy who reported Madoff to the SEC sounds interesting: https://en.wikipedia.org/wiki/No_One_Would_Listen


Agreed, this is a great book that goes over many different types of fraud with amusing examples.


Self plug: If you are interested in short selling and activist funds, I run an online community where people post original research and discuss other short theses: https://activist.cafe/


it would be interesting to codify the known heuristics of forensic accounting into a software package (maybe even AI all the things!), then create a company out of it in partnership with prosecutors and regulators to root out fraud like this through both technological and legal means. i'm sure there have been some efforts to that end, though not commonly known.

it would help rebalance the legal system toward the service of people over corporations.


If you can reliably determine what companies and securities are fraudulent more accurately than the rest of the market and faster than government entities, then you can use this to make rather a lot of money as a short-seller.


Well, not really. As the saying goes, markets can remain irrational a lot longer than you can remain solvent.


You may have problems when the fraudsters run the government.


You're absolutely right. This particular approach has limitations and potential shortcomings. Of course, the same might be said about any approach.

Do you have something else in mind? A discussion here could be very interesting!


something like https://linkurio.us/

built on Neo4J, that's how I know about them.

on edit: you might also like to read https://commons.erau.edu/cgi/viewcontent.cgi?article=1401&co...


That looks strikingly similar to what Palantir’s Gotham product looks like.


The problem is getting anyone to care while the fraud is still making money


that's the point of combining it with lawyers--you find clients, possibly grouped into class actions, to sue the fraudsters (and make money that way).


So an interesting fact about class action lawsuits:

There were initially created to combat civil rights cases. For a long time, minorities who were seeking legal relief against large companies could not afford to mount legal battles.

The class action lawsuit was created as way for plaintiffs to pool resources and incentivize lawyers to take these kinds of cases. Only later on was the class action lawsuit applied to regular civil cases.


Modern Business 101!


Suppose you've made the AI, it's running, and it tells you some company is fraudulent. You're going to be stuck investigating to figure out if it's correct, and if so, what the fraud is exactly.

Which probably isn't that much of a gain. Finding a company that seems shady doesn't appear particularly difficult.




Doesn't work.

"Something went wrong We're sorry. This page failed to Outline."



Must not have worked earlier. It works now :)


Can you detect any shenanigans going on at Tesla?

If so, then this would be a very good stock to short.


Aside from this article (paywalled), I liked learning from some story that certain financial / bank employees are required by the Fed or SEC to go on vacation and not have access to their email or phone for a certain number of contiguous days every year. Is this correct -- was it something to do with the Madoff scandal?

As I understood it, the idea was that to perpetuate a financial fraud, a criminal relies on being present and able to intercept / continue feeding fraudulent information to others and not be discovered. Or maintaining some fraudulent trading position.

The tactic I liked about this was the thinking about what's required to perpetrate a fraud, and make the conditions difficult or impossible for someone to go undetected.

Rather than retroactively finding the fraud and just trying to detect it better when it has already happened.


This was definitely a thing pre-Madoff even, my dad had to do this for two weeks every year as a foreign exchange trader.


The first story is about finding out a company is junk, shorting their stock, then releasing the information. Is that not a form of stock fraud, or at least manipulation, as well?


Isn't it just price discovery?

Nobody would be upset if you bought a company you thought was undervalued and told other people they should buy it too.


If you were an insider, yes. If you are an outsider and especially if you disclose the position, no.


Good investigative reporting into shady business practices good for this world.

However short selling companies before dropping your investigation is a massive conflict of interest. Muddy waters is incentivized to produce a scandalous hit piece, rather than necessarily do fair or accurate reporting. They will make money regardless. The damage their reporting does may make a self-fulfilling prophecy, ensuring they don't get a reputation hit for inaccurate scoops.

I'm surprised the economist didn't even include a nod to the potential manipulation going on here.


You are correct, but really only about the first couple of times he does it, subsequent times people know his M.O. and he either gets a reputation for overblown bullshit, or a reputation for finding actual fraud.


How is that different from posting a positive article after investing in a company? Company press releases have a conflict of interest but they still contain valuable data. In the same way bearish research brings up issues that no one else wants to talk about, readers should expect the source to have a bias since no one is completely objective.


Is it a conflict of interest? Seems to be directly related to the short-seller's interest.


Absolutely. Just look at Pershing Square‘s Bill Ackman. He goes on CNBC essential calling for the end of the world, turns out he had credit protection on bonds and tuned a huge profit[1].

[1] https://markets.businessinsider.com/news/stocks/bill-ackman-...


And a thousand other hedge fund managers talk about how bright the future is after investing in stocks. Maybe we shouldn’t rely on hedge fund managers for our opinion of the economy.


I believe them about as much as I believe the promoters of "Get Rich Through Real Estate" seminars. If their methods worked, they'd be getting much richer doing real estate than pushing seminars.

I recorded one once on TV, and carefully went through the spiel. It was fairly complex, and I was suspicious it was hiding something. Turns out, it relied on tricking the other party into accepting a bond with a maturation value of $10,000 instead of $10,000 now. The money was made on the difference.


> If their methods worked, they'd be getting much richer doing real estate than pushing seminars.

More generally, one should be suspicious of anyone whose business model is not "get rich doing X" but rather "get rich by selling the secret of how to get rich doing X".


Bill Ackman had already sold half his hedge and moved aggressively into stocks when he said that. He talks about it on Farnam Street if anyone is interested.




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: