Quality control: When profits run ahead of cash

Technology industry accruals, however, have been considerably more volatile. As shown in Figure 2, accruals have recently risen well above their historical average relative to other industries.  Higher accruals mean that reported earnings are accompanied by a larger increase in net operating or financial assets relative to cash generation. This may reflect working-capital movements, investment in long-lived operating assets, acquisitions, financial-asset changes or accounting judgements. It can be, but is not necessarily, evidence of poor creditor management or low-quality earnings.

It is important to note that these industry-level measures are capitalisation-weighted. This matters for passive investors because a small number of very large companies can meaningfully influence the industry characteristics within an index. In the case of Technology, several of the largest companies combine high reported profitability with elevated accruals.  High accruals signal a widening gap between profits and cash generation. That can reflect weaker cash conversion and lower earnings quality, but also investment in long-lived assets, including AI infrastructure. As a result, passive investors may inherit these characteristics in their portfolios without making an explicit investment decision to do so.

As Figure 3 shows, the dispersion of accruals within Technology is quite wide. The median company sits much closer to the average across the broader developed-market universe.

The largest-capitalisation stocks are therefore responsible for a substantial share of the industry’s elevated accruals profile. To illustrate this point, we grouped companies into ten accrual deciles and calculated the market-capitalisation weight represented by each decile over time (Figure 4).

The concentration is striking. Within Technology, the highest-accrual decile accounts for more than 60% of total industry weight.

This does not mean that 60% of Technology companies have weak earnings quality. Rather, it means that companies representing more than 60% of the industry’s market capitalisation fall into the highest-accrual decile. That is a concentration effect, not a breadth effect, and it highlights how stock counts and benchmark weights can tell very different stories.

Interestingly, a similar pattern emerged during 2012–2014, when the highest-accrual decile accounted for roughly 50% of industry capitalisation.

What followed was not particularly favourable for the industry from a factor perspective.

Companies with the highest ROA, many of which were Technology stocks, delivered strong performance leading up to 2012. Thereafter, their relative performance weakened. At the same time, companies with low accruals, many outside the Technology sector, were comparatively weak during 2011–2012 but subsequently outperformed during 2012–2014, as shown in Figure 5.

As the old saying goes, “History doesn’t repeat itself, but it often rhymes.” None of this implies that the Technology industry will necessarily underperform in the coming years because of deteriorating Accrual scores. Past performance is not indicative of future results.

It does, however, highlight a potential risk worth monitoring.

The broader economic and market backdrop changed materially during the previous episode. Euro-area tail risks receded, monetary policy supported risk appetite, and investors shifted their focus from crisis resilience towards recovery and future growth. Within Technology, mobile computing, cloud infrastructure and platform business models were displacing the economics of the PC era. Those developments provide plausible context for the divergence observed at the time, but they do not establish causality.

Today is not a replay of that period. Recent one-year factor readings are negative for both profitability and accrual signals, while three-year readings differ only modestly. The lesson from 2012–2014 is a narrow one: during periods of transition, markets can value current profitability and earnings quality differently.

The current AI investment cycle makes this distinction especially relevant. Technology companies are committing substantial capital to infrastructure whose ultimate returns are still being established, even as current profitability remains exceptional. Our evidence does not attribute factor performance to AI. Rather, it shows that profitability and accrual characteristics have diverged during a capital-intensive phase of technological transformation.

The parallel with 2012–2014 is therefore a question rather than an answer: how should investors weigh exceptional current profitability against the capital required to secure future growth?

The differences are important. Investors in 2012 were navigating a post-crisis recovery alongside the transition from PC-centric business models to mobile, cloud and platform ecosystems. Today’s leading AI investors are predominantly highly profitable incumbents rather than emerging challengers, and the monetary environment is very different. The rhyme lies in the tension between current returns and the investment required to secure future growth, not in any prediction that the same return patterns will follow.

The FTSE Russell Quality framework is designed precisely for this type of challenge. It combines ROA, change in asset turnover and accruals into a profitability assessment, which is then complemented by a leverage measure based on operating cash flow relative to debt. When one component reaches an extreme, the others provide context and balance.

When profits run ahead of cash, the key question is not which measure is “correct”. The more important questions are why the two measures disagree, how concentrated the disagreement is and whether the conclusion remains robust under a broader definition of Quality.

The historical comparison is not a timing signal on Technology or AI. It is a reminder that Quality is not a celebration of high profits. It is a test of whether those profits can endure.