Foreword
In 2023, Spain’s non-financial corporations (NFCs) invested 173 billion euros in fixed capital, 5 billion euros (current) less than in 2019. Adjusting for the trend in the prices of capital goods (the deflator for gross fixed capital formation), corporate investment was 30.7 billion euros lower in 2023 than in 2019 (17% lower). If investment had held steady at 2019 levels in constant euros, between 2020 and 2023, the corporate sector would have increased its stock of productive assets by 95 billion euros more than it has in actual fact. The corporate sector’s “non-investment” represents over half of the total NGEU funds. The lag in the recovery in corporate investment in the wake of the pandemic contrasts with the trend in output and employment in the NFC sector, which by 2023 were a comfortable 3.1% and 8.8% above 2019 levels, respectively.

The goal of this paper is to analyse the capitalisation process in the NCF sector in Spain in recent years, putting it in context with the recovery path etched out from the lows of 2012, dragged down by the consequences of the financial crisis of 2008 and its scars, with the ultimate aim of better understanding the reasons for the lag. [1] Our analysis is limited to the NFC sector because of the comparability of the accounting information available and its importance to the economy, as it contributes over 80% of the Spanish private sector’s GDP.

Estimating the stock of productive capital at the NFCs
The NCFs’ stock of productive or fixed capital (their balance sheets also include financial assets and working capital that are not within the scope of this analysis) is the result of a dynamic process shaped by the new investments made by the corporations individually over time and their depreciation as a result of usage, wear and tear, technological obsolescence or other factors. The stock of productive capital at the end of each year t is calculated using the perpetual inventory method using aggregate investment flow and capital consumption figures for the overall NFC sector published in the national accounts. Essentially, this method consists of acknowledging that the fixed assets accumulated by corporations over time are valued at the prices at which they were acquired at the time and that the corporations do not replace their depreciated assets with identical assets but rather replace the productive capacity lost by investing in new capital goods that are generally technically superior to those depreciated. Since the prices of capital goods change over time, the sum of monetary investments made in the past may include monetary units that do not correspond to the equivalent in comparable productive capital services from one period to the next.

The perpetual inventory method calculates the stock of capital at current or replacement prices and subsequently deflates that value to obtain a homogeneous measure of homogeneous capital service units. The calculations use information about the initial stock of capital published by the Bank of Spain and data for the flows of gross fixed capital formation (GFCF) and capital consumption published in the national accounts. The trend in the price of capital services over time is measured using the gross fixed capital formation deflator (GFCFD). Lastly, we assume a rate of embodied technological change of 1%. [2]

Exhibit 1 provides the trend in the stock of NFC fixed productive capital in Spain at current prices and constant 2011 prices between 2011 and 2023. The stock of capital in constant euros increased at a healthy rate until 2019, up a cumulative 18% from 2011. In the years following the pandemic and the subsequent recovery, the stock of capital held stable in time, before shrinking a little in 2023. The stability in the capital stock means that each year corporations have been investing roughly the amount needed to replace, at current prices, their capital as it depreciated, whether through consumption or technological obsolescence.

The fact that the stock of capital in constant euros trended above the stock in current euros for much of the period analysed is primarily attributable to the general downtrend in fixed asset prices, measured using the GFCF deflator, between 2008 and 2014 (Exhibit 2). For comparative purposes, Exhibit 2 includes GDP and the labour cost index. The values for each year are rebased to 2019. Deflation in the prices of capital goods following the financial crisis (the comparable GFCFD in 2008 was 1.31, compared to practically 1 in 2011) did not affect the companies’ final sales prices (GDP deflator) or their labour costs, which were relatively stable. The GFCFD bottomed out in 2014, going on to recover, albeit slowly, until 2019, virtually recovering that year, closing in on the price level of 2011. In 2020, the deflator stabilised and for the rest of the period, until 2023, increased considerably, accumulating growth of 17.5% between 2019 and 2023. Output prices and labour costs also increased during the episode of inflation, by a cumulative 13%, which is less than the increase in the prices of capital goods. [3]

NFCs in Spain: Inputs and output
Capital services are combined with labour services to add value to the intermediate products that corporations purchase from external suppliers. Exhibit 4 shows the trend in the two factors of production and in estimated output in the NFC sector during the period of analysis, rebased to the 2019 values. Output is estimated by deflating the gross value added (GVA) of the NFCs, using the GDP price deflator (Exhibit 2). The labour factor is estimated by deflating the NFCs’ employee compensation costs by the labour cost index (Exhibit 2). The stock of capital is that shown in Exhibit 1.

Until the pandemic, in 2019, the NFCs’ inputs and output move in tandem, with output particularly correlated to the labour factor. The onset of the health crisis in 2020 deviated those trends. The pandemic adversely affected output (decreasing by 14% in 2020 versus 2019) and the labour input, although less so (decrease of 9%). The growth in the stock of capital slowed considerably but remained positive until 2022, probably shaped by inertia around investment projects already in progress that were not cancelled, sufficient to delay the contraction in the capital stock (nevertheless very minor, at -0.4%) until 2023. Exhibit 4 clearly illustrates how the recovery from the worst of the pandemic has changed the mix of inputs used in production, marked by more intense use of labour relative to capital.

Mix of production inputs
Companies that seek to maximise their profits choose the mix of inputs, capital and labour for our purposes, to use in their production process as a function of the cost of the inputs and production technology. Specifically, the ratio of capital to labour used to maximise profits is correlated positively to the ratio of the elasticities of output to the quantities of inputs based on the production function and inversely to the relationship between the unit cost of capital and cost of labour. Therefore, for a given production technology, changes in the relationship between the prices of the inputs will lead to changes in the intensity of capital per unit of labour used in production, marked by more (less) capital per unit of labour when the price of the cost of labour increases (decreases) relative to the price of the cost of capital.

In the years following the pandemic, 2021-2023, the user cost of capital moved away from the cost of labour, implying a relative increase in the cost of capital as an input. Framed by the profit maximisation goal outlined above, the change in the relative prices of these inputs would be expected to lead to more intense use of labour and less intense use of capital in the production process, as is borne out by Exhibit 4. Elsewhere, Exhibit 6 shows how, in general throughout the entire period, the changes in the relative input prices have marked the direction of the change in the capital-labour production factor mix.

The 2023 data point depicts a clearly atypical situation: following the reduction in relative capital/labour costs that year, shaped by the double impact of a slight decrease in the cost of capital and an increase in the cost of labour, the replacement of capital with labour would have been expected to ease. But that did not happen. Perhaps the companies are not yet perceiving the reduction in the financial component of the cost of capital and are continuing to use labour intensely in light of the fact that wage growth remains moderate.

Potential output
Exhibit 4 points to a bigger decrease in output during the pandemic and subsequent recovery than might have been expected as a function of the availability and use of the capital and labour inputs over that same period of time. The shock implied by the health crisis may have altered the relationship between the inputs and output beyond what can be explained by the change in relative prices. To analyse this possibility we perform a comparison between observed output and potential output in the NFC sector (Exhibit 7). Potential output is calculated using a hypothetical production function with two inputs, capital and labour, using the amounts behind the indices provided in Exhibit 4 and assuming regular growth in total factor productivity of 0.5% per annum, accumulating year to year from 2011 on. [4]

Potential and observed output overlap between 2011 and 2019, which means that production technology and the assumption regarding disembodied technological change (cumulative annual rate of 0.5%) provide a true reflection of the productive potential of Spain’s NFCs as a whole. The 2020 pandemic opened up a considerable gap between the two, with potential output trending above actual output.

Return on and cost of capital
The relative prices of the capital and labour inputs explain the trend in the mix of factors of production when seeking profit maximisation for a given technology. Expansion of productive capacity and potential output, however, depends on the economic inventive to invest. Before the pandemic, the growth in the stock of capital of around 2%-3% per annum indicates strong incentives. Those incentives, whether positive or negative, in turn depend on the relationship between the return on and user cost of capital, specifically whether or not the return is higher than the cost.

Conclusions
The recovery in investment in productive capital by the Spanish corporate segment to levels that compensate for the consumption of capital and allow growth in the stock of available capital is important to delivering sustained and quality growth (productivity gains through the technological change embodied in capital goods and disembodied technological change). This article demonstrates that the stock of capital was sensitive to the trend in the relative prices of capital and labour throughout the period under analysis, except for an anomalous outcome in 2023, which is when the financial incentives to add to the stock of capital were insufficient (the return on capital equalled its cost). For the corporations to have the incentive to increase their stock of capital at net positive rates similar to those observed in the years prior to the pandemic, the return on those investments needs to comfortably outstrip their user costs of capital. In other words, the return on capital, still below pre-pandemic levels in 2023, needs to increase, with inflation in capital goods returning to 2015-2017 levels. Nevertheless, the comparison between potential and observed output in recent years suggests that the pandemic alone may have destroyed up to 25 billion euros of output in constant euros, equivalent to roughly 5% of the NCFs’ output in 2020.

The picture painted in this paper of the accumulation of productive capital in Spain is partial. Not only because it leaves out the financial institutions, unincorporated businesses, households and the government, but also because the capital assets of the universe of NFCs within the scope of this analysis are aggregated, without considering their composition. The national accounts include investment in R&D and intellectual property in gross capital formation but leave out other intangible assets (such as organisational capital). Moreover, within tangible assets, the gross capital formation aggregate does not distinguish between investment in housing (which commands a significant weight on the NFC balance sheets published by the INE) and investment in other assets such as capital goods or transport equipment. Not to mention the loss of accuracy implied by working with aggregate sector data rather than individual company data. Nevertheless, within the body of work that can be considered comparable in terms of methods and data used, we believe this analysis provides new and relevant perspective on the possible reasons for the lag in the recovery in productive investment in Spain in the wake of the pandemic. Essentially, there has been a lack of financial incentives to add to the stock of capital: despite considerable growth in nominal profits, that growth has been insufficient to offset the parallel increase in the user cost of capital.