- 23 minutes ago
Category
🗞
NewsTranscript
00:00Shares of IBM trading near their lowest level since November 24, September of 2024.
00:05The company dialed back its full-year sales forecast after a pretty steep drop
00:09in mainframe sales, which weighed on results.
00:12Joining us from New York is the co-host of Bloomberg's The Close,
00:16Ramain Bostic, alongside IBM CEO Arvind Krishna. Ramain.
00:20Arvind, you've seen the reaction amongst investors here.
00:23Some concerns here about that lowered sales forecast overall,
00:26as well as softness in software.
00:28You've characterized this as basically a shortfall for one quarter
00:32that is limited to CapEx-sensitive areas of the portfolio,
00:36but that's still a meaningful area of your portfolio.
00:39Were sales coming into that quarter? Was that pipeline overstated?
00:44I don't believe so, because when we look at all the deals that didn't close,
00:49I think we have done enough verification, including with the clients,
00:53to know that they were very real, and it was a reprioritization,
00:58of the CapEx spend at the end of the quarter.
01:01This was pretty confined to, I'll call it the Fortune 100,
01:05the deals that were within a subset of those.
01:08Now, one-third of what didn't happen has already come back.
01:12So that tells us that this was a reprioritization,
01:15and those deals were very real,
01:16as opposed to us being optimistic in our projections.
01:20And, Romain, I would also add,
01:23I think that maintaining our free cash flow tells us that we have levers around productivity
01:28and conviction and confidence in the business,
01:31and that is also, I think, going to serve our investors well.
01:34But the next few months will tell us that.
01:36Well, on that cash flow figure, yes,
01:38and that certainly pleased a lot of analysts and investors out there,
01:41that billion-dollar number of projected free cash flow growth,
01:44you're maintaining the dividend as well.
01:46But you've largely done that so far by cutting costs.
01:49So that raises the question that if we are anticipating slower growth on the revenue side,
01:55does that mean more cost cuts are in store?
01:59So the bulk of our cash flow growth over the last four years
02:03has actually been on adjusted EBITDA.
02:06So that tells you that this is mostly through revenue growth,
02:09and our model has always been that the last dollar is more productive
02:14and more profitable than the first dollar.
02:16So we've been growing revenue 4-5%,
02:19and we've been growing cash flow up in the 7-8-9%.
02:22So that's kind of our model, and we intend to keep maintaining that.
02:26Now, right now, if we drop revenue by one point,
02:30because we said 4-5 instead of 5-plus,
02:32we can absolutely make it through productivity.
02:36Cost cuts is an interesting question.
02:39Cost cuts doesn't always come down to people reduction in headcount.
02:44Our headcount has been more or less flat over the last many years.
02:47I think there's a lot of third-party spend
02:49where we are going to get a lot more efficient
02:52with that third-party spend than we have been always.
02:56Bloomberg Tech is live on Bloomberg Television and Bloomberg Radio,
03:00and we're speaking with the IBM CEO, Arvind Krishna.
03:02Arvind, good morning.
03:04You want to focus on accelerating revenue growth
03:07and accelerating profitability.
03:09And just really simply, I'd love to hear what you're asking the team
03:12to do differently now in response to all of the factors that you outlined.
03:17So, Ed, really, so if I look at our software business,
03:2280% of it is already an annuity, consumption, OPEX-based business.
03:2720% of it is a CapEx business.
03:30If we think that the CapEx headwinds are going to continue,
03:34but 80% is already growing at about 8%,
03:37we want to put a lot more focus,
03:39so we are going to direct a lot of the team
03:41with forward-deployed engineers,
03:43with people who are focused on deploying the software at clients,
03:47much more technical help,
03:48and make that 80% grow even faster.
03:51Products like Red Hat, Confluent, Hashi,
03:54all fit that model.
03:56On the CapEx side,
03:58we have to make sure that while we can continue to do it,
04:00don't depend upon outsized growth on that side to go there.
04:04Then, on the supply chain,
04:07can we leverage all of our capability in the supply chain
04:10to make sure we have enough distributed infrastructure
04:13in storage, in Unix systems,
04:16that people can fulfill all of the demand?
04:18Because we came out of the second quarter
04:20with half a billion dollars of backlog
04:22in that part of the portfolio.
04:24So, those give you an idea of the kind of changes
04:26that we are making already,
04:28not just for the rest of the half.
04:31You summarized the state of the world beautifully.
04:34You said that customers shifted spending
04:37towards servers, storage, and memory in late June.
04:41And when I posted on social media
04:43you were coming on the show,
04:44the question from the audience is really simple.
04:46Did that trend continue from June into July?
04:49And for how long do you expect it to last?
04:53We have not seen it in July,
04:55but I would tell you, look,
04:57the semiconductor pricing,
04:59memory is up, what, three to four times
05:01over the last 18 months.
05:03Networking infrastructure is up 60 to 80 percent.
05:08Fiber is up.
05:09Connectors are up.
05:10I do think that some of these trends
05:12on the underlying components
05:13are going to carry on for some more time.
05:16That works its way up into those things.
05:18So, people are going to have those prices.
05:21Now, the question is,
05:23is that going to maintain a reprioritization
05:25of the capex spend?
05:26If I take the signal
05:28that a third of our deals closed,
05:29then that says, no,
05:30people are going to get more careful
05:32about how they spend their capex.
05:34If that goes into September,
05:36because a lot of capex does get committed
05:38at the end of a quarter,
05:40not always through,
05:41then it means that that carries on
05:44for some more time.
05:45And that is why we gave a guide
05:47of four to five percent,
05:49depending upon if that carries on,
05:51we'll be at the low end of the range.
05:52But if some of it comes back,
05:54then we'll be at the high end of the range.
05:55So, Arvind, I know there's a big focus right now
05:57on sort of the clients
06:00that sort of did not necessarily materialize
06:02in the quarter.
06:03With regards to the existing clients that you have,
06:05and I'm primarily referring
06:06to your mainframe business,
06:08I mean, can you share
06:09like sort of what percentage
06:10of those clients are actually
06:11increasing their spending?
06:13Is that going up still?
06:14Absolutely.
06:15So the current machine is called the Z17.
06:18The Z17 has been from the beginning,
06:22which was May of 2025 to today,
06:25at 130 percent in terms
06:28of the capacity growth
06:29compared to the prior machine.
06:30That's one very strong signal,
06:32and that's an aggregate.
06:34Then you can say,
06:35hey, is everybody in there
06:36or only a few?
06:3785 percent of the clients
06:39are increasing their capacity
06:41as opposed to the 15 who are not.
06:43That optimization goes on all the time,
06:46and I will tell you
06:46this is probably the best
06:48that we have seen in a long time
06:50of what is going on there.
06:52The software is going to lag.
06:54Yeah, software is going to lag.
06:56I mean, look,
06:56we know the hardware capacity is there.
06:57Software is lagging,
06:58and I understand why
06:59it's lagging in the moment.
07:01And I'm going to ask you a question,
07:02and forgive me if it's a bit unfair,
07:04but I look back to the 1990s
07:06and whether there are some parallels
07:07to some of the mainframe issues
07:09that IBM went through back then
07:11under a predecessor
07:12two or three times removed from you,
07:14John Akers,
07:14and this idea that at that time
07:17we were going through
07:18this paradigm shift in computing
07:20and how companies spent
07:21and allocated their money.
07:23And there was a lot of talk
07:24by the CEO then
07:25that the issues were temporary,
07:27that they were economic,
07:28and that they were rights themselves.
07:29We know in hindsight
07:30that it was much more structural.
07:31Why should we now look at this
07:34major shift going on with AI
07:36and the computation involved in that
07:38and think that this time is different?
07:42Well, we have to look at
07:43what the clients are doing.
07:44I always start there.
07:45Our opinion is an opinion.
07:47What clients do is what matters.
07:49So when I check with my clients,
07:51those who do credit card authorizations,
07:53are you going to maintain the mainframe?
07:55And they go to, well, the resilience,
07:57the amount of capacity,
07:59the unit cost of it being 5 to 15 times cheaper
08:03is important.
08:04The big difference from that time,
08:07early 1990s to today, Romain,
08:09is the cost issue.
08:11You could not argue
08:13that the mainframe was cheaper
08:15on a unit cost basis
08:16for the workloads that are moving off
08:18than at that time,
08:19the mid-range computers.
08:21I'll call it the whole Unix.
08:23It wasn't really client server.
08:24It was Unix taking the workloads off
08:27in the early 1990s.
08:29That issue you have to look at.
08:31If there is a cheaper alternate
08:33for that workload,
08:35I'll sort of look at you and say,
08:37that means in 5 to 15 years,
08:38it will move off.
08:39But that's not the case.
08:41Right now, for the workloads that are on,
08:43we can show the clients
08:44that it's 5 to 15 times cheaper
08:46to keep it on the mainframe.
08:47They're not.
08:48But that's not all workloads.
08:49That is workloads that need protection,
08:52that need resilience,
08:53that need the burst capacity that comes.
08:56And so for those kinds of workloads,
08:58the mainframe is the architecturally
09:00superior platform.
09:01That was not the case 30 years ago
09:03for the workloads that did go off.
09:07Live on Bloomberg Television
09:08and on Bloomberg Radio,
09:10this is Bloomberg Tech
09:11speaking with IBM CEO Arvind Krishna.
09:13I've been listening to you a lot
09:14recently on long-form podcasts,
09:17some clips on the social media
09:18about your view of the world,
09:20what it's like to be a CEO
09:21in the domains that IBM operates in.
09:24And it's interesting how quickly
09:26it comes back
09:27to the macroeconomic backdrop.
09:28A lot of people looked at the guidance
09:30for the balance of this year
09:32and would say,
09:33and they do say, Arvind,
09:34on this show,
09:35memory prices are not changing.
09:37They continue to push higher.
09:39They look at the outlook for growth.
09:44Could you just explain
09:45the data points you rely on
09:46that give you the confidence
09:48on the new guidance that you've given
09:50and whether it is actually achievable
09:52or it will be difficult to meet?
09:54Yeah.
09:55So, Ed, I'll come back to,
09:58for ourselves,
09:59we have to look at our demand pipelines
10:01and our yields
10:02and how we get things going.
10:04But I think for your audience,
10:06let's look at it this way.
10:08Number one, most important,
10:09is what is GDP growth going to be?
10:11We think that that's between
10:132% and 3% for the year
10:14if I look at the globe.
10:16And it is going to be consistent
10:18even in the Middle East,
10:19even in Asia,
10:20where there is a lot more energy
10:22and disruption,
10:23we actually see a lot of growth.
10:25Tech is going to be,
10:26I think,
10:272 to 3 to 4 points above that.
10:29So that puts tech
10:31in terms of what the market is
10:32somewhere in the 5%, 6%, 7%.
10:35Then that comes back to
10:36what parts of our portfolio
10:38can play against that demand
10:39and what parts cannot.
10:41And so I look at the parts
10:43of the software portfolio,
10:44that's why I talked about
10:44the 8% growth
10:45in the 80% of it,
10:47that can play right into that.
10:49Then I look at
10:50our distributed infrastructure,
10:51that can play right into that.
10:53I think consulting
10:54will be in that 1, 2, 3%.
10:57It is not going to be
10:58in the double-digit growers.
11:00But we see the demand
11:01and we see the signings
11:03and we see the clients
11:04leaning in to say
11:05they want transformational work done.
11:07That's how we kind of know
11:09beginning with the macro
11:11and then coming down.
11:12I'll make a prediction for you.
11:15I think technology spend
11:16is going to become
11:17a larger and larger part
11:18of every enterprise's budget.
11:21It used to be 3%
11:22since Romain raised
11:24the 30 years ago.
11:25It's probably up
11:27at 5, 6% on average.
11:28I will not be surprised
11:30if by 2035,
11:30it's 10% of everyone's budget.
11:34Give you a quick
11:35micro case study.
11:36Bloomberg reported
11:37that Starbucks is replacing
11:38some IBM tools
11:39on the software side
11:40with in-house.
11:42Talk to that.
11:44So Starbucks is about
11:45a little over $2 million
11:47a year client for IBM.
11:49The portion they're replacing
11:50is a product called TriRiga
11:52that does real estate
11:53lease management.
11:55The version of it
11:56that Starbucks has
11:57is almost 10 years old.
11:59I'm not surprised
12:00that they're replacing it
12:01because I've actually been
12:03describing publicly
12:05software which is largely
12:06interaction based
12:08and is based on ease of use
12:10as opposed to anything else
12:11can be easily replaced
12:13by AI and AI agents
12:15and that is what is going on there.
12:17However,
12:18if I see other parts
12:19of Starbucks
12:20and maybe the ability
12:21for our Harshi portfolio
12:23or security portfolio,
12:25how about if I phrase it this way?
12:27I would not be surprised
12:28if Starbucks
12:29is a larger client next year
12:30than it was last year.
12:32With regards to how
12:33this world is evolving,
12:35Arvind,
12:35I am curious just internally
12:37about your plans to hire,
12:39particularly when it comes
12:40to the technological side.
12:43Have you been able
12:44to keep pace
12:45with some of the other
12:46companies out there
12:47also trying to do
12:48what you do
12:49and pay some of the salaries
12:51that you have to pay?
12:53Well,
12:54we hired three times
12:55as many college hires
12:57this year
12:57than we did last year
12:59and I think that
13:01given others seem
13:02to be backing off
13:02college hiring,
13:03it gives us
13:04an incredible ability
13:05to bring in great talent
13:07and to then grow them
13:09inside our company
13:10and to offer them
13:11great careers.
13:12I think if I looked
13:13at it last,
13:14we had I think
13:1620 million resumes
13:17in our applicant database
13:19so that gives us
13:20a huge field
13:22to go look at.
13:23So, I mean,
13:24I don't worry
13:25about bringing in talent.
13:27I actually worry
13:28much more about
13:29can we give them
13:29a great career
13:30and a great pathway
13:31because not everybody
13:33is cut out
13:34to do work
13:35that is going
13:36to be demanded
13:36because I don't think
13:37there's a lack
13:38of employment
13:39but the nature
13:40of the work,
13:40if you can't use
13:41AI tools,
13:42if you can't use
13:44the productivity tools,
13:45then it's going
13:45to be really hard
13:46for you to be competitive
13:47with your peers.
13:49We already know
13:50sort of the potential
13:51impact of what AI
13:53means for the economy
13:54and for your business.
13:55There are a lot of people
13:56looking around the corner
13:57including yourself
13:58to quantum.
13:59Is that a viable business
14:01on the horizon
14:02or is that just
14:03a moonshot
14:03that you're hoping
14:04actually sticks?
14:06Well,
14:07when things are
14:07two years away,
14:08I wouldn't call them
14:08a moonshot.
14:09I think quantum
14:10is now in the
14:11engineering realm
14:12as opposed to
14:13the science realm
14:14for the next
14:15five years
14:16and I believe
14:17that by 2029
14:18we will deliver
14:20a machine
14:20that does
14:21100 million
14:22computations
14:23of large-scale
14:25fault-tolerant
14:26quantum computer
14:27in the next
14:28two years.
14:28I think that's
14:29an incredible
14:30opportunity.
14:31I'll quantify it
14:32and this is not
14:33just our work,
14:34a lot of third parties.
14:34by the end
14:36of 2030s
14:38we think it's
14:39about a trillion
14:39dollar total
14:40market opportunity
14:41in terms of value
14:42that quantum
14:43will create.
14:44That is why
14:45we also doubled
14:45down this morning
14:47and announced
14:47that we bought
14:48HRL
14:49from GM
14:51and Boeing
14:51and that's
14:52going to help
14:53us bring people
14:54with a lot of
14:54talent around
14:55materials,
14:58spintronics,
14:59quantum sensing
15:00and other
15:01sensor technologies
15:02to add to
15:03our own effort
15:04so we are
15:05even more
15:05well positioned
15:06to go win
15:07in this market.
Comments