The Rise of Cloud Computing: Minding Your P's, Q's and K's

terrific new Corrado, using AWS data finds massive falls in prices not being captured by GDP

after his adjustment"....boost the average annual growth rate during 2007-2015 by roughly 2 percentage points compared with official estimates. For nominal GDP growth, including this own-account investment would add three basis points per year to the growth rate during this period..."

David Byrne, Carol Corrado, Daniel E. Sichel, "The Rise of Cloud Computing: Minding Your P's, Q's and K's," National Bureau of Economic Research, October 2018, https://www.nber.org/papers/w25188

"....Cloud computing—computing done on an off-site network of resources accessed through the Internet—is revolutionizing how computing services are used. However, because cloud is so new and it largely is an intermediate input to other industries, it is difficult to track in the U.S. statistical system. Moreover, there is a paucity of systematic information on the prices of cloud services. To begin filling this gap, this paper does three things. First, we define the different segments of cloud computing and document its explosive expansion. Second, we develop new hedonic prices indexes for cloud services based on quarterly data for compute, database, and storage services offered by Amazon Web Services (AWS) from 2009 to 2016.....Our basic finding is that prices for cloud services have fallen rapidly and that the use of the cloud has grown tremendously as has investment in the related infrastructure of IT equipment and software. For our analysis of prices, we assembled a unique data set with quarterly data on prices and characteristics for cloud services offered by the largest provider, Amazon Web Services (AWS), since the first quarter of 2009 when AWS began posting prices on the Internet. The data cover AWS’ basic compute, database, and storage products..... These indexes fall rapidly over the sample period, with quickening (and double digit) rates of decline for all three products starting at the beginning of 2014..... For AWS, the price for their compute product fell at an average rate of about 7 percent during 2000-2016. Price declines were slower before 2014 and more rapid starting in the beginning of 2014. Interestingly, 2014 is the year when Microsoft and Google began posting prices for their cloud offerings on the Internet. We suspect that AWS’ large price declines were a response to that change in the competitive environment. For AWS’ database product, prices fell at an average rate of more than 11 percent during 2009-16. Here too, prices fell relatively modestly until the beginning of 2014, after which they fell at an average rate of more that 22 percent through the end of 2016. AWS’ storage product followed a similar pattern, with prices falling at an average annual rate of about 17 percent during 2009-2016 and even faster declines starting in 2014.....Finally, we highlight the puzzle of why investment in IT equipment in the NIPAs has been so weak while capital expenditures have exploded for IT equipment associated with cloud infrastructure. We suggest that cloud service providers are undertaking large amounts of own-account investment in IT equipment and that some of this investment may not be captured in GDP..... As seen in figure 9, nominal IT equipment and software investment including our estimate of own-account would be $58 billion higher in 2015 than in the official estimates, amounting to 0.32 percent of GDP. For nominal investment in IT equipment, adding this own-account investment would boost the average annual growth rate during 2007-2015 by roughly 2 percentage points compared with official estimates. For nominal GDP growth, including this own-account investment would add three basis points per year to the growth rate during this period..... We find that cloud computing has exploded. By available measures, the quantity of cloud activity has grown extremely rapidly as has associated capital investment. At the same time, prices of basic cloud services have fallen rapidly since 2009, based on a unique dataset we assembled. However, because cloud is so new and so much of it is intermediate input, it is challenging to track in the statistical system, and the available data do not distinguish between cloud-based and traditional services, whether services are purchased or produced internally, or generated at the “edge.” We highlight one area where real GDP may be understated by a noticeable amount as a result of changes in the economy related to the rise of cloud computing...."

would note that while it might be missing in GDP Mr. Market seems to be pricing it into AMZN/MSFT value versus peer firms not in the space