Much has been written about the growth of debt since the Great Financial Crisis. But whilst leverage is a worry, much less has been written about the problem of asset quality
As someone writing about derivatives as the last crisis hit in 2007 (or 2008 depending on your markers), I always try to keep track of those markets which have both the opacity and scale to become systemically destabilising. People are already staking their claims on what will be the next ‘CDO crisis’, whether its sub-prime car loan asset-backed security markets (which are not large enough to cause systemic problems, in my opinion) or exchange-traded funds (which probably are). But there is a possibility the next crisis may start somewhere unexpected: within the non-financial sector; or at least interactions between the non-financial and financial sector.
One of the most interesting developments in many large cap non-financial firms is the growth of intangible assets, in particular goodwill. In simple terms, goodwill is what an acquiring firm records on its balance sheet when it pays more for a target company than the value of that firm’s book price. The growth of goodwill owes much to changes in international accounting rules just over a decade ago. Prior to 2001 under FASB and 2004 under IFRS, it was a requirement that goodwill be amortised (depreciated) over a maximum of 40 years. This changed as the influence of neo-classical economics and a shareholder value conception of the firm began to shape accounting standards. It was argued that because goodwill is an asset like any other, it should be valued on a net present value basis, representing the market’s best estimate of the discounted future cashflows that arise from intangible things like brands, consumer loyalty etc. Because firms have indefinite lives, and because there is no replacement cost for goodwill as there is for machinery, amortising goodwill distorts useful information for investors about a firm’s ‘true’ value. Furthermore, it was argued that ratios like Earnings Per Share or Return On Capital Employed were already affected by the liabilities used to finance an acquisition, so the additional requirement to amortise goodwill was a form of unfair double counting. The accounting rules FAS142 and IAS36 abolished amortisation, and instead made goodwill subject to an annual fair value assessment, which means goodwill is only impaired when its fair value is deemed to be less than its carrying value.
Like many changes built upon economic abstractions – the purism of the logic breaks down upon contact with the realeconomik of market practice. In the real world, goodwill is not an asset like any other, it is simply an accounting artefact to deal with a double entry problem; and cannot be seized by creditors when a firm goes bust. Markets also misprice things, and that includes corporate assets: many M&A deals fail, so there is no ineluctable relation between goodwill and future cashflows. Firms do not have indefinite lives: the lesson of Enron or Carillion is that the more you believe that they do, the more irresponsible risks you tend to take. And finally, the benefits of goodwill as an asset do expire over time (brands do fade) and so it makes sense to factor this cost into any prospective acquisition price and the costs of capital taken on as a consequence. The more telling risk is that there is a form of double counting that arises when goodwill is not amortised…
Here’s an alternative story about the last 10 years since the crisis. If we start by sidestepping the metaphysical question about whether firms have indefinite lives and engage with the reality that the median tenure for a large-cap CEO is 5 years, it changes the way we think about goodwill. The most obvious point is that over the tenure of the average S&P500 CEO, goodwill acts as a different form of double counting: when a firm is acquired, the future cashflows from the purchased entity are capitalised on the balance sheet as goodwill. But the actual cashflows from that acquisition are then recorded each year without any corresponding reduction in goodwill. Because future income has already been capitalised once, this gives CEOs the option of building cash mountains (capitalising income twice) or putting earnings to some other use. Faced with the prospect of investing in projects which are uncertain and where gains may be realised outside their (short) tenure, CEOs have tended to distribute; and they distribute in part because it triggers bonus payments for themselves but also because (paradoxically) hitting these shareholder value targets is the easiest way of ensuring their tenure is extended.
What we are seeing in large cap firms is therefore an attempt by corporate elites to impose their own temporal preferences for income today onto the firms that they manage. This results in practices like levering up against speculative assets such as goodwill to bring forward income from which dividends can be paid (NB goodwill is speculative because its value reflects expectations about future cashflows and discount rates). It will also mean there is a reluctance to impair goodwill to try to minimise costs and maximise returns in the present. This is a financialized practice: as I’ve argued here and here – financialized capitalism does not just denote the squeezing of labour costs to pay out to shareholders. Neither is it completely captured by shifts in the financial sources of accumulation. The financialization of the firm is expressed through the migration of financial logics across the economy. This means that ‘going concern’ strategic calculations are increasingly displaced by an approach which treats the firm as a space for materialising extractable funds created by a kind of yield spread: between the risk free income that can be brought forward and crystalised and the present costs of future liabilities, which can be passed on, substituted or minimised. Financialized capitalism is, in short, about staggering the temporalities of asset-based income and liability-based costs to produce a yield in the present.
In a context where cheap credit is abundant, the double-counting magic of goodwill accounting has encouraged CEOs to pay increasing multiples of annual earnings for corporate assets: S&P500 firms are currently paying a median of 12.4x EBITDA for their acquisitions – higher than before the crisis, dwarfing the 11x earnings that were, for example, deemed to be reckless by the UK Treasury Select Committee’s investigation into private equity in 2007. By removing the obligation to amortise, acquisitive firms have become less price sensitive – if goodwill was expensed to earnings, CEOs would have to be more circumspect about the prices they paid over book for corporate assets because amortisation charges would erode the distributable funds from which dividends could be paid.
This is resulting in significant changes in the balance sheet profile of corporate America. Figure 1 shows a stacked distribution of intangibles share of total assets in a stable panel of S&P500 firms between 2008-2017. The panel is comprised of any company which spent any time in the S&P500 between those dates (including dropouts) and for which data on intangibles and total assets are available for all years. In this dataset I removed new entrants and inactive firms, leaving 462 firms in total. This panel is likely to underestimate the growth of intangibles because it includes dropouts who have been either less acquisitive or whose underperformance may have led to goodwill impairments. It also excludes a number of more intangible new technology entrants who entered the index after 2008.
Figure 1: Intangible assets as a % of total assets, S&P500 members by quartile 2008-2017
Plot generated by Mark Taylor, Sheffield Methods Institute (@markrt)
The data shows there is a lower quartile of firms whose intangibles share of total assets is structurally fixed at below 10% for all years. This includes industrial asset-intensive stalwarts like Ford who have limited opportunities for large scale M&A which might alter its goodwill profile, or Wall Street banks like JP Morgan and Morgan Stanley whose assets one might think were ‘intangibles’ but are not classified as such. Outside that quartile, intangibles as a proportion of total assets have been climbing. This shift is more evident if we create a sub-panel by only including those companies who have registered an intangibles to total assets ratio of 20% or above at one point in the years 2008-2017 (Figure 2). This reasonably isolates those firms who don’t have activity requirements for large amounts of tangible assets and thus have some exposure to what I have termed ‘futurity risk’ – the risk that future cashflows and discount rates which items like goodwill embody do not materialise as expected. Within this sub-panel, we can see a clear increase in the number of firms with intangibles as a percentage of total assets more generally; and we can see a particular increase in the ratio of intangibles to total assets in the third and fourth quartiles between 2015-17.
Figure 2: Intangible assets as a % of total assets, sub-panel of S&P500 members (intangibles: total assets >20%) by quartile 2008-2017
Plot generated by Mark Taylor, Sheffield Methods Institute (@markrt)
The above figures report intangibles more broadly, but much of the growth of intangibles has come from goodwill. According to Bloomberg, between 2008-2017 the lump of goodwill in the S&P500 rose from $1.8 trillion to $2.9 trillion. That additional $1.1trn is large enough to be systemic. If we follow the logic that balance sheets are becoming more like a trading asset portfolios, non-financial firms are effectively financing speculative, intangible assets like goodwill with the fixed contractual obligations of debt. When thinly capitalised entities have a levered exposure to speculative assets, that is cause for concern: shareholder equity as a proportion of total liabilities is on average lower now than in 2008 when many S&P500 firms had just charged losses to their equity buffers.
For current goodwill valuations to hold, the future needs to be stable and predictable. Firms are carrying more goodwill than ever before at a time when the world is becoming anything but. As one hedge funder recently noted, the global economy is exiting ‘the long season of central bankers’ and entering the ‘lush, wild tangle’ of politics. If growth stalls or interest rates rise this could trigger the reverse of the double counting phenomena: firms will have to restructure current operations, resulting in large exceptional items which put pressure on their equity buffers. But they should also simultaneously impair their goodwill to reflect the diminished future cashflows arising from those restructurings. This could lead to large writedowns at precisely the point that firms’ operating performance is least able to accommodate them. Add leverage to that mix, and there may be inadequate equity redundancies to soak up the losses; or firms will be forced to recapitalise through share issue at precisely the point that the market doesn’t want to buy shares.
This dynamic is eerily familiar to those of us researching finance before 2008 – this is the problem of pro-cyclicality. Indeed, there is a certain irony that the accounting changes designed to improve information for investors may introduce balance sheet volatility which few investors are able to understand or predict. There are already strong incentives to avoid goodwill impairments. Should we experience another downturn, or should interest rates begin to rise, it is likely that we see more corporate failures where balance sheet volatility and accounting scandals go hand in hand, as firms who cannot afford to make goodwill impairments use the opacity and discretion of CGU accounting to avoid them. The risk is – like the CDOs of 2007/8 – that investors simply lose faith in all firms’ representations of their goodwill valuations. That is when we may see a series of containable problems become a systemic concern.