$1.7 billion market cap
$4.2 billion EV
Debt to EBITDA of 1.3X
Massive depreciation and amortization costs of around $1 billion per year.
DXC was formed out of the merger of 2 old school IT outsourcing companies: computer sciences corporation (CSC) one of the oldest IT outsourcing companies founded in 1959; and Electronic Data Systems, which was another IT outsourcing company founded by Ross Perot in 1962. EDS changed hands a number of times before being spun off by HP Enterprises in April 2017.
A lot of their legacy business is running on premise servers and doing direct IT consulting. This is a pretty crummy business nowadays with the hyperscalers offering cloud services and SaaS software decreasing the need for IT consultants.
AI is a double edged sword. It should probably accelerate the decline in the need for IT consultants, but might allow DXC to operate with fewer employees. So maybe revenues shrink but costs shrink faster. (I think this might be the case for a lot of consulting companies but hasn’t really played out yet).
AI might also increase the attractiveness of on-premise servers. So that way you can run some cheap open source LLM on your own server with your own data and no chance of leaking the data to anyone else.
When you hear complaints about how banks run on old COBOL software, they are talking about DXC’s Hogan software, which was written in COBOL in the 1970s. It’s really difficult to displace this software since it’s hard to have downtime on people accessing their deposits, they have to maintain meticulous records and regulatory compliance, and the costs of any data loss are very high. It looks like banks are moving towards keeping Hogan and then put in APIs that can talk to it using more modern software.
DXC has had this legacy asset sitting around but it makes up a relatively small piece of the overall revenue, maybe a few hundred million out of several billion. But it is pretty high margin. It looks like their latest plan is to jack up the costs for banks and switch to more usage based pricing, so they can exploit the high switching costs. This should add some more high margin revenue and stem the decline in profits.
The revenue has declined every year for the past 8 years and guidance is for another year of decline. Revenue isn’t forecast to stabilize until 2029. So it’s a melting ice cube.
Trailing operating cash flow is $1.5 billion, but includes a $200 million litigation settlement.So if we take that out we get $1.3 billion of operating cash flow, with $230 million of capex and $200 million or so of software development costs. So I get to something like $800-900 million of trailing FCF per year. This puts the company at a P/FCF of 1.8-2.1.
Going forward that cash flow might continue to decline but you get some offset if they can really mild Hogan for some extra cash flow and some optionality if AI drives some more on-premise server demand.
They seem to be using the cash flow to pay back debt and buy back stock. Diluted shares are down from 255 million in 2022 to 174 million today, and down 3.4% YoY in latest quarter. They paid down $500 million of debt in the past 12 months.
Anyway thought this was kind of interesting.