The IT Sector Could Weather The Pandemic Storm
April 6, 2020 Timothy Prickett Morgan
It has been a rough couple of weeks for absorbing exponential data. It is astounding has fast the Great Infection, my term for the combination of the coronavirus outbreak and the reprise of the Great Recession that it looks like it is causing, is upon us. Nearly 10 million people have lost their jobs in two weeks, and my guess is that will more than double again next week and keep building from there, perhaps quadruple until we go from 3.5 percent unemployment in the United States to maybe 15 percent or so.
The hope is that this is temporary, and that we can rebound quickly once the coronavirus pandemic crests and abates. The Chinese manufacturing powerhouse, which was shut down by the pandemic a month ago, is starting to come back online, and it will probably be months before the United States is back to work. What has us concerned is that the coronavirus pandemic might return in the fall and maybe even have another spike in the early spring in 2021. This triple hit happened with the H1N1 Spanish flu pandemic of 1918 and 1919, as you can see here. And it kept going into 1920. Depending on who you ask, about quarter of the population of Earth got the Spanish flu, which was nicknamed that because Spain did not fight in World War I and its newspapers were not subject to censorship and reported the spread of the disease, and somewhere between 3 percent and 5 percent of the population – about 1.8 billion – in those years died. Depending on the assumptions, anywhere from 17 million to 100 million people died. The consensus was around 40 million.
Here is our point: We were not idiots a century ago. We kept records just as poorly as we are doing now and we had the same reactions of disbelief and then acceptance as the disease took root and ripped through city and country alike. It was extremely disruptive, and it is certainly not a coincidence that the second – and worst – wave of the Spanish flu coincided with the end of the War and the signing of the Treaty of Versailles. The Spanish flu had reached its highest of three peaks in New York, Paris, and Berlin in late October 1918 and in London in early November, and the war was declared over on November 11, Armistice Day.
Many experts, including Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases at the U.S. National Institutes of Health, think we will have a second wave of the novel coronavirus, and hopefully not a more virulent strain. Fauci said last week that it was “highly likely” that we will see the coronavirus infections return. This time, we will have months to manufacture all of the right stuff and get organized if we have to shut the global economy down by region one more time until we can find a vaccine or get herd immunity.
Given all of this, it may seem premature to make predictions about how the economy will rebound in the United States and elsewhere, and how spending in the IT sector – including spending on hardware, software, and services as well as payroll for all manner of IT employees, which is a much, much larger number – will react to all of this. But this is where we work, and let’s face it, we are all wondering this. The experts at IDC have given a first pass on this, which is interesting, and it is of course based on assumptions that may not be valid depending on how much of a shock the global economic system just succumbed to. The point is, knowing what we know now, the short term is certainly awful, but the experts think that at least as far as the IT business is concerned, the hit is not as bad as many might expect.
IDC has been adjusting down its IT spending projections (which are measured in constant currency) as the coronavirus outbreak gathered steam, as you can see in this chart below, which shows global IT spending increases:
When I look at this chart, my brain calls the Probable IT spending growth rate of around 3.75 percent “March projection” and the Pessimistic IT spending growth rate of what looks like 1.3 percent “August projection” and my eye just wants to take that all the way down to close to zero in one called “December projection.” But that is now what IDC’s economists and market researchers are saying, despite the tricks my eyes are playing on me.
Here is the same data forecast from 2019 through 2023:
IDC says IT spending growth was up 5 percent in 2019, and that it would increase to slightly more than 5 percent growth in 2020 before the pandemic hit, and then hover just shy of 5 percent out in 2021, 2022, and 2023. After the pandemic, IDC is still expecting growth this year, and then some makeup growth in 2021 as we come out of this – in the current Probable scenario, the decline this year is nearly made up by supplemental increases in 2021, and in the Pessimistic scenario, the incremental growth in 2021 is not enough to make up all of the lost revenue in 2020, but at least some of it is covered. The lesson here is that the more difficult this gets, the harder it will be to get any kind of rebound to the normal IT spending growth – GDP plus a few points – we see on a global basis. But there are no negative numbers in any of these scenarios.
To be fair, this is IT wares spending, not IT payrolls, which will be a different story. We have no sense of what is happening out there, so you can enlighten us. We suspect that few companies are going to furlough their IT departments for one or two months, even if they shut down retail, distribution, and manufacturing operations, because of the critical nature of IT in the modern business. But we have no data on this, only gut.
Here is what IDC’s projections look like by the type of spending for 2020, based on the pre-coronavirus and the probable and pessimistic scenarios after the pandemic broke out:
Mobile phone sales are somewhat insulated and software sales – thanks in part to the subscription rather than perpetual license sale of system and application software these days – look pretty insulated. Ditto for IT services, which will see more modest growth. PCs and tablets were going to take a big hit after an aggressive refresh cycle in 2019, and servers were not going to grow much either and even in the pessimistic scenario, the revenue decline is only a few points.
From the look of things, however, non-X86 systems, such as IBM’s System z and Power Systems iron, is going to take more of a hit. Here are scenarios that IDC has cooked up for server sales in 2020:
The 2019 figures for non-X86 iron were buoyed up by System z15 sales in Q3 and Q4 last year, so that is a tough compare going forward. But it looks like in the current climate, after a 16 percent decline in sales for non-X86 iron (which includes servers based on Arm chips, on the rise ever so slowly in the datacenter) this part of the market is going to stabilize somewhere around $7 billion a year. X86-based machines will take a hit, we think mostly due to enterprise, government, and educational institution spending cuts for servers, and the hyperscalers, cloud builders and service providers who make up a little less than half of the market will fill in most of that gap created by the other industry sectors. From there on out, the X86 market recovers, according to IDC, and somewhere around 2023 X86 iron will break through $100 billion in sales.
The idea encapsulated in these numbers is that the Great Infection created an artificial recession that is more or less easily reversible provided there is enough economic stimulus. So far, governments seem pretty eager to print money and what we really need to worry about is the combination of high unemployment and inflation. But when you are in survival mode, as we are now, you can’t worry about that. Survival is the first order of business.
We think it is highly unlikely that in the current climate that IBM i shops are going to be eager to do much spending on systems, except maybe cloud capacity for high availability and disaster recovery. And even that may be put off for now. We do expect for companies to try to keep their application modernization efforts on schedule. And of course, those who are in the middle of replacing systems because they need capacity no matter what are going to finish those upgrades – particularly if they think the budgets for them might be rescinded if they don’t spend it quickly.