To open with Microchip's recent leadership change, which has seen you return as the interim CEO, and announcements you made yesterday, that you know, revenues are at the low end of guidance and that Fab 2 will be closed. So I just wanted to open it up and see if you wanted to say anything about that.
Sure, so let me sort of start with the leadership change. I wanna start by, you know, thanking Ganesh Moorthy, who worked for Microchip for 23 years, and you know, he was a very hard worker, had a number of positions at Microchip for five years, from 2016 to 2021. He was the president when I was CEO, and he was two in the box with me. I couldn't have had a better wingman. You know, he was just a great executor when I pointed him in the right direction, and then in 2021, I stepped down from the CEO role, and he became the CEO. The results, especially in this cycle and the last couple of years, Microchip has substantially underperformed, so he stepped down from his role and retired, and you know, I had retired completely. I actually had moved out of the Valley.
I still had a house here, but I was living in Prescott in the mountains, you know, having fun, hiking peaks, navigating lakes, and playing with the grandkids and family. And then the Board, you know, asked me if I would step back in. And, you know, I with my team built this company for 31 years, and when the duty called, I couldn't say no. So, you know, here I am. I moved back in the house in the last week. Two questions people have is, you know, what does interim mean? My title is Interim CEO and President. Number two, what am I going to do different? The definition of interim is I'll be in this role for as long as it is necessary. You know, it is not weeks. It's much longer.
You know, Microchip is, you know, if you look at our performance today, you know, we're down 55% in revenue from the peak, 1,000 basis points down in gross margin, 2,000 basis points down in operating profit, severe cash crunch. We're not producing even enough cash to pay our dividend. We're borrowing money to pay dividends. So I kinda look at it as a turnaround story almost. And we need to get all these things fixed before there is even a thought of another CEO transition. So there's no search out today. When I'm looking for another CEO, you know, I'm it for a while, whether you like me or not. So the question should I keep going? I have questions.
Yes.
Okay.
Keep going.
I keep going. I just didn't wanna, you know, have you get cut off from your question. So what am I going to do? You know, I'm on a, you know, a massive urgent but sense of urgency program to review entire Microchip from top to bottom in the next 90 days. And I'm gonna give you nine points planned that we're gonna look at. The first one is resizing the manufacturing footprint. So Microchip's manufacturing footprint today is too large. We had a peak revenue, which was more than 2x of our current revenue, and we were adding capacity to grow beyond that. And some of that capacity is installed, and much of that capacity is not even installed. It's an equipment sitting in our fabs that we can install and go substantially above our prior peak.
So at the current revenue level and with any kinda expected growth, even, even significant growth, we got inventory, just incredible amount of, you know, incredible amount of capacity. And in trying to keep those fabs with some reasonable utilization, we have continued to build inventory, and we just can't do that anymore. We're drowning. So, yesterday, we decided to close down one fab, you know, Fab 2, which is our smallest fab. And a lot of people have been asking the question that by closing that fab, are we cutting off our upside? I'm really not worried about the upside. We have so much capacity and so much inventory. It was the smallest fab. The other fabs are capable of doing three times of what they're doing today. So, so the upside is not an issue. So that's number one.
And so this is resizing just the fab, but we have other factories. We have four factories, large factories in the back end, two in Thailand, two in Philippines. And we have other, you know, smaller fabs around the world, two in Germany. We have a plant in Ireland. We got three in Massachusetts. So all that structure, I'm gonna look at it and review, you know, what needs to be combined, what needs to be done, what. So the second issue is the inventory, which is tied to the capacity. For 31 years, I ran Microchip as a CEO. I never had inventory go higher than 135 days on our balance sheet. Last quarter, the inventory in our balance sheet was 246 days. And this quarter was gonna go to 276 days. And that was before we lowered the guidance yesterday. So it's gonna go higher than that.
The outside inventory, what we buy from foundries, is about 234 days. So the inside inventory is even higher than the number that I gave you. It's probably closer to 300 or higher. That's, you know, over 2x the inventory. I mean, you know, it's just like way too much. Enough is enough. So we're going to aggressively bring the inventory down. Our inventory is too high, even on the foundries. So we're dramatically cutting starts from the foundry to bring that inventory down. The system inventory is closer to 300 days. Systems is 10% of our business, where we build boards and, you know, systems and development tools and others. Why does it have 300 days of inventory? It should be 100 days lower. So I'm, you know, gonna mount a program to bring that down.
The third one is, you know, there are a lot of growth initiatives in the company. There is Megatrends. There is TSS. There is AI. I need to review each one of those programs and see what is working, what is not working, maybe take one away, maybe add one, you know, add more resources to this because there's higher potential, take resources from this because it's not working as well. So I'm gonna do that entire review in the next 90 days. Then the fourth one is a review of the business unit. We have 25 business units in the company. We actually, every quarter, see 25 P&Ls. Some of the business units are quite small. They were started for a reason, maybe five years ago, 10 years ago. Is that reason still valid? Is that promise still there?
In some cases, maybe it's not. So they should be combined or closed or sold, or I'm just going to review everything and see what needs to be changed. So that's the fourth. The fifth one is a channel strategy. So you're aware of significant changes in the distribution strategy in our industry, where, you know, many other, you know, companies have done. Microchip channel strategy hasn't changed in several years. We do business with about 100 distributors around the world. Some are global, some are regional, some are catalog. And I'm going to do a comprehensive review of that channel strategy to see what are we paying everybody. You know, is that too much or is that too little? Should something be terminated? Should something be added? And I think the whole channel strategy needs to be looked at again. Number six is strengthening our customer relationships.
I think in the times of extreme shortage and after that, when the industry came down and there were a large number of non-cancelable, non-reschedulable orders on our books, and in certain cases, customers didn't want the product, but since we helped them in good times, in bad times, we wanted them to help us and keep taking that product, I think we stressed certain relationships and, you know, call it an apology tour, call it a forgiveness tour. I think we are on it, and, you know, we got 125,000 customers. We cannot visit them all, but, you know, we need to take top 5,000 customers and ensure that, you know, we love them, but do they still love us? And some of them have six months of inventory. They're not buying as much from us today because they have inventory.
But when the inventory runs out, are they still our customers? They're gonna still buy from us, and we have the next design. I think we do. I think it's okay, but we just need to fall on the sword and apologize and, you know, give them some TLC and ensure that our customer relationships are strengthened. Number seven is a business model. Now, you know, our business model that you have still has remnants of elements from peak performance. When, you know, growth was very high, gross margin was very high, there were expedite charges, there were price increases, customers willing to do anything to get the parts. And the operating expense percentage was very low because the revenue was so high, and you couldn't add operating expenses fast enough.
So the business model that is still in, you know, in our presentation, which has been taken out, really had remnants of peak performance in a supercycle, which is not representative of the real world. So I'm going to recast that in 90 days. I don't know what changes will happen in there. When I look at on the gross margin side, actually, some of our businesses that are higher than corporate gross margin are doing better than our businesses that are lower than corporate gross margin. So gross margin could actually be accretive, for all I know. But there is also pricing pressure, things happening in China.
So there are things on both sides of the ledger, and I'm gonna need to put it all, process it all together, and see what the new growth number is, what the new gross margin is, operating expenses, and give that business model to you and then aggressively execute to get there. So that was number seven. Number eight is on the OpEx side of the equation. Our OpEx today is, you know, 35% of our revenue. And that was before we lowered the revenue yesterday, so it's even higher. Plus, all year, we have been on a pay cut. That pay cut was reversed two Fridays ago, just before I joined. It was the decision made before. So when I look at the next quarter, the full pay will be for the entire quarter. And when that happens there's some water in there.
So when we have a full quarter of full pay, the OpEx in dollars will rise some. Plus, in the first quarter, all the FICA comes back for the higher-paid people because usually in the third and fourth quarter, the company matching part of FICA is not there. So I think I'm concerned that the operating expense goes higher and gets farther away from a model that I like to see at. So what needs to be done? What portion of it, you know, will actually self-correct when the revenue comes back, and what portion doesn't come back, and we need to do something in terms of weeding and feeding and business units or whatever? So that's just a question right now. I don't have the answer to that. And the final one, number nine, is, review the CHIPS grant.
The grant was applied to maybe almost a year ago when everybody thought that the factory capacity was never enough and the world is gonna build silicon fabs forever. And, you know, today, we have too much capacity. And, you know, government gives you 15 cents of every dollar you spend. Do I wanna spend $100 million to get $15 million back from the government? If I don't need it, then I don't wanna spend the other 85 myself. So, you know, I don't know what the answer would be. You know, with 10 days back at the job, I just need to reopen that. I have put the negotiations with CHIPS office on hold for now until I get my arms around it. And probably by the time I get my arms around it, we're into the new administration.
I don't know what their take is, which people they put. Is the money still available? But whatever happens, then we will deal with it. So that's kinda what I'm going to do. And plan is to, you know, get this accomplished in 90 days, which is a Herculean task. You know, I don't know how much of it gets done, but that's what the attempt would be. And we'll bring some of the output back to you probably at one of the conferences in March, and give you some updates. So that's kinda where I am. Sorry for the long answer.
That's great. That's great. Thank you, Steve. As I hear you talk, and I think about how sharp the recovery could be for Microchip in particular, given just the depth of how far things have come down, I would think that when things come back, they come back pretty sharply and more sharply for you than maybe some of your peers. Now, that might be a while from now. We can debate when that is. But the pace of the recovery ought to be pretty quick because of your reliance on distribution and, you know, all those factors. So how much does that come into your mind when you're thinking about these cuts? You don't wanna overcut. So how do you? I mean, you know, you've been around a long time. You've seen a lot of these cycles.
So, you know, I'll address the second part of your question regarding the cut, and then maybe Eric can answer how sharp the recovery can be. You know, I'm not worried about the upside portion of it. You know, I take the example of if you ask me to hit the bullseye, you know, chance for me to hit the bullseye is zero. But if you say, "Make sure that you're on the left of it and never go right," then I can do it every time. I'm not worried about the right side of it, the upside of it, because we have so much inventory and so much capacity. Even after closing this fab, our other two fabs can do 3x of what they're producing today. So I have no concern about the upside.
I have the concern on the downside because, you know, for four or five quarters, we've been calling it a bottom, and we're probably still, you know, dribbling at the bottom. The current quarter is down what, 12% from last quarter sequentially? The inventory has been growing. Every quarter, we guide down, and the inventory grows further. I think that's my concern. The upside is covered. We just wanna make sure that through all these actions, we don't build inventory. You know, today, there is probably $800 million to $1 billion of excess cash tied up in inventory and CapEx. If we didn't have that, we probably didn't need to refinance one of our loans.
One of the things we need to do is liberate the cash through inventory, liberate some cash through CapEx, and liberate some cash through growth of the business that is likely to come in the coming quarters. By combination of all those, get the free cash flow above where our dividend is. Today, we're borrowing money to pay the dividend. There's no plan to cut the dividend. Dividend is not at risk. We may have to borrow money for a couple of quarters, which we will do. But as we liberate cash and the cash flow becomes higher than the dividend, then we'll take the remaining upside cash and pay back down this loan that we're borrowing right now to pay the dividend.
Once we have done that and the total debt has gone back to where it was pre-borrowing for dividend and we have more money left over, then we'll have a conversation about what we do with it. But that's not, you know, immediately in the next quarter or so, so I'm not worried about the upside, but, like, you know, what sort of recovery you think can happen? What do you wanna say about that?
We know that there is a lot of inventory that is being drained throughout our customer base and their customer base today. I mean, we can quantify the piece that's distribution. I mean, we've quantified the last three quarters that spent about $100 million per quarter that distribution sell-out has been in excess of what we're selling in. We expect something around that again this quarter. That is with distribution sell-through falling because distribution customers are draining inventory, right? That's on a little less than 50% of the business that goes through distribution. Same thing is happening on the direct side, but we don't get any kinda real-time reporting from the customers.
I'm with you that I think this can have a really nice bounce off the bottom, but projecting where exactly that bounce takes us to is an impossible thing to do today. We have very limited visibility.
Sure. Got it. And, Steve, we do all these scrapes of disti websites, and microcontrollers in particular are sort of an outlier in terms of the inventory sitting at these distis. And it's not just Microchip. Is there something unique about MCU? And, I mean, part of it probably was PSP and all the things that you talked about. Yes, there is that, you know, lag effect where you were putting product to customers. But, but how much of sort of what's going on in microcontroller is industry-related? And, you know, how would you tie that back to the environment for demand across a different market?
I don't know if there's anything in a microcontroller product compared to any other product that it should have a higher inventory. You know, I can't put my finger on it, but, you know, one thing I look at is, you know, TI, Silicon Lab, Microchip, NXP, they all had new CEOs in the last business cycle, and they all built a huge amount of microcontroller inventory. I don't know if there is anything else common. I think for all those CEOs, it was the first business cycle. You know, it just I don't know how else to say it.
Okay. And then can you talk to China? I know this is more of a structural question. People, you know, China's sort of seen as this melting ice cube, and it's seen as just only a matter of time until everyone in the industry begins to feel the indigenization of what China's trying to do. How much do you see from your seat? How much do you see the indigenization efforts as a threat to your business in China?
So historically, you know, what we have said is, you know, China is 18% of our business, and half of that business is selling to multinationals who are doing manufacturing in China. But that product then moves back out to U.S. or Europe or Japan somewhere. And, you know, we didn't think that that one was at risk because products they're manufacturing for Western will not likely use a local, you know, Chinese parts. Then the other 9% of the business that's remaining, you know, half of that are very complex parts: FPGA, data center parts, you know, high-end Ethernet parts, high-end microcontrollers, you know, well-integrated parts, so which are not the parts that Chinese are trying to make or copy and all this boutique Chinese industry developing.
So if you take all that out, it only leaves about 4% or 5% of the business, which is commodity-like microcontrollers that are going to the Chinese customers, which could be vulnerable to the local Chinese supply. So that's really what we have historically said. And it's largely still correct. But I think you're starting to see some chinks in the armor where, you know, for example, European car manufacturers have assembly plants for cars in China, and they're getting government pressure to use local Chinese parts. And, you know, here and there, you're starting to see designs pop up. It's not making any significant difference. It'd probably be, you know, years before it makes any meaningful difference. And automotive design is three years away from production, usually. It's a long cycle. But I think, you know, that thing is no longer zero.
There are some cracks starting to develop, and then China is applying more and more pressure for their industry to use local parts. Now, local companies don't really have a broad enough product line. You know, I have said for years, if you recall this conversation, that usually 70% of microcontroller customers do not go to production with the device they first designed with. You know, a year or a year and a half, it takes to go to production. Usually, the spec changes. Marketing comes in and says, "Well, we need to add a color display. Black and white is not good enough." You know, USB connectivity was fine, but now we need to connect it to internet. Or the power is too high. We need a low-power part. You know, so during that design phase, the spec changes.
When that spec changes, if you're designing with a local Chinese person who has two or three microcontrollers, you don't find another part that fits in there which allows you to do what you do. We have 20,000 microcontrollers, distinct microcontrollers, parts with USB, Ethernet, Wi-Fi, Bluetooth, PWMs, you know, low power, high power, high speed, low speed, pin counts, 8-pin, 16-pin, 32-pin, 40-pin, 68-pin. So when the customer definition changes and then he needs a different part, he can preserve his investment in design and firmware and usually find a different part which has what it needs from a Microchip portfolio. He can't find that kinda diversity to fit into the same socket from a local Chinese guy who designed two parts.
And the second thing is when a Chinese guy designs one or two parts, there is no microcontroller in our portfolio that will make even 0.1% of our business, the 20,000 mask sets that make up our business. So it's highly fragmented, and that high fragmentation is our friend, you know, because you can't attack that kind of fragmentation. You have to produce all these devices. So competition is increasing. And you know, but hundreds and hundreds of these Chinese companies funded by government are going bankrupt in two, three years. But some of them are prevailing. And the ones who are prevailing are, you know, starting to make some small inroads, no matter how small they are. They're not affecting the revenue today. You know, it's of the order of, you know, less than 1%. But it's an issue in the next five years and 10 years.
So what is needed is really and some of our competitors are looking at the same thing, where you need a China-for-China strategy, you know, where, you know, there's a Chinese company that we fund, which is based in Shanghai with a Chinese name, with a Chinese logo. We have actually just trademarked a Chinese Microchip. You know, it's when you read it, you know, it's, it's in Chinese, but apparently it says Microchip. I can't tell, and then, you know, you sell the die to this company, and this company locally assembles and tests and markets it with a Chinese data sheet, with Chinese engineers, with Chinese management, with a Chinese managing director. You know, that's one model. There are versions of that other model. People are talking about it.
So once you become that, you know, then you're part of the ecosystem with the strength of our 20,000 parts, with the strength of our design capability, marketing muscle, you know, all the app notes and, you know, website and on and on and on. So you take your global strength and you apply it with a Chinese frontage. And then the locals cannot make a dent. And you participate in the growth of the Chinese business. We're gonna land up doing something. I don't know what that form would be. So I think that's the answer there.
Yeah. I mean, I think some of the challenge, though, is that the rules are ever-changing. When it works today, it might not work 12 months from now, right? So you have to be on top of it and not be so dedicated to one approach and then have that approach not work.
Is there any, as you go through this evaluation, who knows what the new administration will do, but the talk about tariffs and all this kinda stuff, I mean, how does that play into your thinking about what you wanna do?
So, you know, what the tariffs and all that have done is, you know, it has stressed relationships with Chinese customers. You know, I was talking to our head of applications in China a couple of years ago. And he was complaining that some Chinese customers are saying, you know, "U.S. government shut down ZTE. They shut down Huawei. You know, why should we design with an American product? They could shut us down someday." So, you know, so, you know, I said, "Well, let's take the example of ZT, ZTE. They sold parts to Iran in violation of the embargo. Don't you think they should be penalized for that?" And his answer was, "Why does China care? That's U.S. problem. You don't like Iran. They are our friends." So that's the attitude, you know, and that's the attitude.
You know, they say, "Hey, we don't have any problem with Iran," so the problem is, you know, when we are shutting down those customers, how would our customers in the U.S. feel if China shuts down Cisco over there, like we shut down Huawei? You know, when you think of the reverse of that, it's frightening. But one thing China hasn't done so far is they haven't retaliated, you know, much at all. I think there was some sort of note this morning that they're gonna control germanium or something. But all this time, while we have put sanctions on them and we have put tariffs on them, they haven't retaliated.
If Trump comes in and puts more sanctions and shuts down more companies and things like that happen and the relationship got even worse, and then Chinese start to retaliate, hold rare earth materials, you know, not allow us to sell and all that, then that becomes much more difficult. But my take on that is, number one, it's not likely to happen because China and U.S. cannot exist without each other. There's too much dependence on materials and products and others. Neither can Chinese nor U.S. can complete a server or equipment without having some material or parts coming from the other country. So I don't really think the self-destruction is likely to happen, you know, so, you know, some wisdom will prevail there.
Secondly, you know, whatever happens, whatever rules are, you know, our, our job is to achieve what is achievable better than our competitors. And if you can do that, then that's the best that can be achieved. Today, we are underperforming to our competitors, and that's not acceptable. And that's. I'm here to change that. And, you know, whatever happens, if under those circumstances, under those rules, think of it, you know, I cannot change the flow of the river, but I can navigate the boat around the boulders and around the waterfalls and, you know, achieve the maximum speed, and get ahead of the competitors. That's what we need to do.
Great. Eric and Steve, thank you so much.
Appreciate it.
Thanks.