A recent series of announcements from the new Trump administration indicate that changes in tax policy, including proposed incentives for corporate capital investment, are on the horizon.
An earlier proposal included deductions for 100% of asset acquisition and production costs -- in effect allowing companies to expense capital investment immediately, rather than through a multi-year depreciation cycle. (See Trump & Consequences.)
While the intent of the changes is to stimulate infrastructure and manufacturing spending, it may tempt CFOs and CIOs to revert back to on-premises technology spending -- a remarkable reversal. At CEB, our budget benchmarks (full report for members only) research has shown a continuing decline in average capital expenditure among corporate IT organizations, with zero growth in capital expenditure expected for 2017.
The culprit behind this trend is the cloud: This year, a quarter of IT organizations will allocate 20% or more of total IT spending to cloud providers. (Full report for members only.)
However, will this reversal actually happen?
The temptation will be real.
CEB data shows that, even if cloud unit costs -- and opex spending -- wind up being less expensive than equivalent on-premises costs, the prospect of IT costs savings remains a long-term bet. IT organizations migrating systems to the cloud typically see a significant "year-one bump" in opex that pushes cloud's break-even point several years down the road -- if at all.
From a financial standpoint, the promise of savings several years in the future may pale next to the potential tax savings of capital spending in 2017.
Even with the temptation of saving money this year, CFOs and CIOs should still carefully consider pulling back from the cloud for these four reasons.
1. Capital investment should support digital ambitions, not data center ambitions: It's myopic to limit the "capex vs. opex" discussion to "the cloud versus the data center." For the majority of Fortune 500 enterprises today, technology is essential to enterprise growth ambitions. Companies announcing large-scale shifts to the public cloud, like GE and Johnson & Johnson, don't do so in isolation, but in the context of supporting digital strategy and revenue goals. As one vendor's tagline puts it, cloud may be less about saving money, and more about making money. From that perspective, channeling capital spending to the data center may not make sense if that capital could be allocated more profitably to growing digital business lines. This will matter more if interest rates rise, and if the Trump administration sees through on its promise to eliminate deductions on interest expense.
2. Today's tax savings could be tomorrow's legacy headache: In most IT organizations, support for legacy systems and technologies is the chief resource constraint on innovation, on speed and on most things digital. Capital spent in 2017 should come with explicit communication to business partners about the opex tail that will accrue across multiple years and the difficulty associated with shedding it.
3. It could – inadvertently – bring back the question of 'why does IT cost so much?': Even where cloud computing is more expensive than on-premises options, it tends to offer much more straightforward cost transparency to business partners. Organizations furthest into the cloud tend to report that it changes IT's cost conversation with business partners, and moves it away from wasteful, nickel-and-dime tracking. There should be some question of whether it's worth discussing "why IT costs so much," in the absence of a streamlined, mature budgeting and financial process at the IT-business interface.
4. It's possible to capitalize cloud migration and use tax incentives to speed movement to the cloud: Moving to the cloud at scale requires a lot: application refactoring, possible expansions in network bandwidth, vendor negotiations, training, and additional factors. So much so that the "year one" increase in opex can be daunting. However, migration may invite the purchase of new assets -- network equipment, monitoring software, and so on -- that can represent capital expense. Enterprises committed to cloud migration should consider where they can pair their cloud strategy with capital expenditure that facilitates migration, and offers some tax relief on migration expense.
Over the last year, IT leaders have come to recognize that cloud economics is more complex than an "opex versus capex" debate. They should not let corporate tax code changes lead the enterprise to forget this, or overlook the lessons of past IT cost conversations.