John D. Rockefeller and his slick Big City lawyers are credited with inventing the concept of the subsidiary corporation in the 1870s to evade antitrust laws and other regulations in the US oil industry.
Subsidiary corporations have proven a headache for regulators ever since.
By the 1920s, most large US corporations were using subsidiaries for a host of reasons: To facilitate growth and branding, to limit corporate liability, to lower/evade taxes and to assist financing due to the fragmented US banking industry. Power companies, in particular, relied heavily on subsidiary companies to fund the expansion of the electricity grid.
In the roaring 1920s, most companies also used subsidiaries to manipulate their accounting. Although consolidated accounting had been around for centuries, the rules on how to consolidate accounts were loose.
Virtually all large companies used subsidiaries to hide liabilities and expenses. For companies that were less than 50% owned, there were no formal rules. The 1920s were the wild west of accounting.
In 1927 the infamous Harvard economist, Professor W. Z. Ripley, devoted a large section of his book, Main Street and Wall Street, to warn about the problem of subsidiaries and over-complicated financial accounts.
The Harvard Law Review followed with a stinging criticism of credit manipulation in subsidiaries in 1928. When the crash came in 1929, companies were found out. Many companies had more debt, hidden in subsidiaries, than their accounts indicated.
The rules for consolidating accounts were progressively tightened. In 1934, the American Institute of Accountants (AIA) issued its first official guidance on the equity method, used for minority-owned investments of between 20% and 50%.
Although the rules for accounting for subsidiaries have undoubtedly improved, the fact is that many companies still use subsidiaries, whether wholly owned or partly owned associated companies, to house liabilities or incur expenses that the parent company does not want to disclose.
Mostly these days it occurs through investments in associates, where the equity share is less than 50%. Rupert Murdoch once famously boated that he could control any company with 22.5% ownership.
The most important way that subsidiaries assist account manipulation is by making the accounts difficult for investors to understand.
Complexity is the key. This is achieved, for the most part, using a witch’s brew of moving parts not related to the core operations of a business.
It is achieved by acquiring businesses, creating subsidiaries and joint ventures, utilizing off-balance sheet vehicles, and having associated companies in far-flung places where auditors are unlikely to go.
With many acquisitions and investments in associated companies, add further complexity with one-off transactions, restatement of subsidiary earnings, non-operating income, write-ups, write-downs, goodwill adjustments, capitalized expenses and complicated intercompany transactions.
Put it all in the pot and keep stirring.
The quality of an audit always suffers when a company’s accounts are muddied by hundreds of subsidiaries and investments in associates.
Auditing is a competitive, expensive and complex business. Audit fees are only loosely based on the complexity of a company’s accounts. The quality of an audit always suffers when a company’s accounts are muddied by hundreds of subsidiaries and investments in associates.
Putting some flesh on this idea, let’s look an example. Alibaba's audit & audit related fees in FY2017 were US$8.2 million. Its audit was undertaken by PwC in Hong Kong, where auditor salaries are similar to those in the UK. In 2017, the average audit fee of a FTSE 350 company was US$5.5 million.
However, Alibaba was far bigger and vastly more complicated than your average FTSE 350 company. Alibaba’s company records are kept in China, not Hong Kong. Moreover, Alibaba operated a vast network of subsidiaries and related entities.
According to its 20-F annual report filing, in 2017 alone, the company added 300 subsidiaries and other consolidated entities.
The unconsolidated investments, of which there were many, were not listed. Alibaba had so many consolidated entities that it only provided an approximate number in its 20-F filing.
It is difficult to believe that any auditor could vet so many new subsidiaries, let alone the existing subsidiaries, for the fee that PwC was charging. This claim was substantiated by this year’s inaugural regulatory audit inspections of Chinese companies by the Public Company Accounting Oversight Board (PCAOB).
In its first round of inspections, the PCAOB investigated four audits undertaken by PwC in Hong Kong and four undertaken by KPMG’s affiliate in China. It found that seven of the eight audits fell short in vetting their clients’ accounting or internal controls. The violation rate was five times higher than that typically found in audits by the Big Four.
We’re not saying that Alibaba was manipulating its accounts. We are simply using Alibaba to demonstrate that audit expenses do not equate to account complexity. The more complex the accounts, the less comprehensive an audit will be.
We’re also not saying that companies with complex accounts are necessarily manipulators. This is an illustration that things can go undetected or are easily missed in such companies. Ill-intentioned companies with complex accounts have better opportunities to manipulate than those without.
When assessing account manipulation risk, our software checks to see whether a company’s accounts are abnormally complex for its size and particular industry. One obvious risk indicator is the number of subsidiaries. How many subsidiaries are normal? How many is too many?
In 2022, the 6,186 largest public companies in the world had an average of 59 subsidiaries. The number of subsidiaries varies according to the size of a company, its industry and the geographic extent of its operations.
Some industries, such as construction and pharmaceuticals, entail more risk than others. Companies in these industries will tend to have a high number of subsidiaries in relation to revenue. Similarly, companies in innovative industries tend to have more subsidiaries and JVs than those in less innovative industries.
In general, company size is the biggest determinant of the number of subsidiaries. Most small- and mid-sized companies have fewer than 50 subsidiaries. Only very large companies tend to have more than 250 subsidiaries.
Figure I: Distribution of companies according to number of subsidiaries
Source: GlobalData
Only a handful of companies have more than 1,000 subsidiaries. Figure II shows the 10 companies with the most subsidiaries. French construction company Vinci wins first prize with a staggering 2,689 subsidiaries.
Figure II: Distribution of companies according to number of subsidiaries
Source: GlobalData
Unsurprisingly, all of the companies in this top 10 list raise concerns on the Transparently.AI account manipulation risk software. Virtually all warrant either extreme caution or high caution in signals of asset quality.
Subsidiaries, especially if you have hundreds of them, would make it easy to exploit loopholes in Generally Accepted Accounting Principles (GAAP). Examples would include non-commenced leases or variable leases.
These are both effective ways to keep debt off the parent’s balance sheet. The impact on funding costs of lessening recorded debt can be large, especially when credit spreads are wide.
Subsidiaries, especially if you have hundreds of them, would make it easy to exploit loopholes.
These types of leases should be documented in the notes to the accounts of the subsidiary - but who will ever see them if a company has more than 500 subsidiaries.