MEC 2026年第三季度业绩电话会：数据中心销售额增长65%

核心要点

• 受数据中心与关键电力、商用车以及建筑与高空作业设备需求拉动，2026年第三季度净销售额同比增长18.1%至1.704亿美元。
• 调整后EBITDA从1410万美元增至1700万美元。调整后EBITDA利润率达到10.0%，高于去年同期的9.8%，环比提升约190个基点。
• 数据中心与关键电力业务销售额增长约65%，约占季度总销售额的22%。MEC斩获了约4500万美元的新订单，合格储备管线规模已超过2亿美元。
• 约2000万美元的高利润率数据中心与关键电力业务收入从2026年下半年递延至2027年。管理层预计第四季度调整后EBITDA利润率将略低于第三季度，随后在2027年实现逐季环比改善。
• MEC更新了2026年业绩预期，预计净销售额为6.30亿美元至6.40亿美元，调整后EBITDA为4800万美元至5200万美元，自由现金流消耗为600万美元至1600万美元。
• 在爬坡成本下降、产能利用率提升、员工队伍扩充以及数据中心业务占比提高的支撑下，管理层继续以约15%乃至更高的调整后EBITDA利润率为目标。

核心财务数据

业务与运营表现

数据中心与关键电力仍是MEC的主要增长驱动力。销售额同比增长约65%，占总销售额的约22%。随着更多项目实现全面量产，产能利用率有所提高，但公司也产生了约300万美元的投产、外包及溢价用工成本。

MEC本季度斩获了约4500万美元的新增数据中心与关键电力订单。这些项目包括发电设备外壳、开关设备、配电盘外壳以及不间断电源（UPS）。在过去一年中，公司在该市场已斩获超过1.45亿美元的项目，且合格机会储备管线超过2亿美元。

在客户需求走强及北美8级重卡产量提升的支撑下，商用车业务销售额增长近35%。MEC还在为客户应对2027年美国环保署（EPA）法规变更的新品投产做准备，预计将于2026年底投产并在2027年期间爬坡放量。

随着非住宅建筑需求改善，建筑与高空作业设备销售额增长约9%。相比之下，受需求疲软及主机厂供应链调整影响，动力运动装备销售额下降约27%。管理层预计该压力将持续至2027年。

农业装备业务仍处于周期性低谷，而军工项目继续面临过渡期延迟。农业装备和动力运动装备业务占比较高的工厂产能利用不足，令利润率承压。

MEC正在其工厂网络中重新调配生产，使项目与可用劳动力和产能相匹配。公司工厂目前有约230至250个职位空缺，其中焊工的招聘尤为困难且成本高昂。新设备将于第四季度末到货，预计将于2027年第一季度投入运营，届时外包生产将逐步转回厂内自产。

公司于10月1日完成了对佐治亚州汤姆森一家制造工厂的收购。包括收购价格、厂房改造及设备在内，MEC预计到2028年将对该基地投资约2500万美元。该工厂预计到2028年将带来约5000万美元的年营收产能，相关投资评估均以不低于15%的内部收益率作为基准。

管理层业绩指引

对于2026年全年，MEC目前预计：

该业绩预期基于全年在财务上并入Accu-Fab、交叉销售带来的增量收入，以及商用车和建筑与高空作业设备市场的持续复苏。这些利好因素预计将被农业装备和动力运动装备的疲软，以及持续产生的数据中心投产爬坡、外包和溢价用工成本部分抵消。

管理层表示，由于客户排期变动，约2000万美元的数据中心与关键电力收入从2026年下半年推迟至2027年。加上新增的人工和投产成本，导致调整后EBITDA指引中值从5600万美元下调至5000万美元。

受季节性因素、客户交付推迟及持续的爬坡成本影响，第四季度调整后EBITDA利润率预计将略低于第三季度的10.0%。管理层预计第四季度将是利润率筑底的基期。

对于2027年，MEC预计利润率将呈现逐季环比改善，而非一步到位。爬坡成本预计将在年初开始放缓，而劳动力可用性的改善、固定成本摊薄以及经营杠杆效应预计将在2027年晚些时候及2028年更加凸显。正式的2027年业绩指引计划在第四季度财报电话会议上发布。

MEC预计2027年资本支出为4000万美元至5000万美元，主要用于支持数据中心与关键电力业务的增长。管理层还预计将战略性退出2000万至3000万美元的低优先级工业项目，从而为更高利润率的数据中心业务腾出产能。

风险与关注重点

• 客户排期变动依然是收入和利润率波动的一大诱因。多家数据中心客户在应对自身供应链瓶颈的同时，将项目订单推迟到了2027年。
• 劳动力短缺推高了溢价工资、加班费以及临时用工成本。MEC预计，随着招聘和内部培训的推进，这些压力将逐步缓解。
• 设备调试以及将外包生产转回厂内自产可能需要一定时间，这或将限制2027年第一季度初期的利润率改善空间。
• 农业装备与动力运动装备的疲软导致MEC部分生产厂区的产能未能充分利用。管理层预计农业装备短期内不会出现拐点，对动力运动装备的销量也持谨慎态度。
• 受营运资金需求和增长型资本支出影响，预计2026年自由现金流仍将为负。
• 数据中心建设正面临围绕税收优惠、电力基础设施、电力成本及水资源利用的政策争议。尽管如此，管理层对未来三至五年的需求前景依然看好，并表示MEC的产品通常在项目开工约三年后进场交付。

分析师问答亮点

利润率走势：管理层确认，预计2026年第四季度将是利润率的底部。新设备预计将于2027年第一季度投入运营，随后外包和溢价用工成本将有所下降。数据中心业务收入占比的提升应会在下半年为利润率的进一步改善提供支撑。

数据中心收入递延的影响：MEC将EBITDA指引下调的大部分原因归结于约2000万美元的数据中心与关键电力收入推迟至2027年。管理层指出，这部分收入的调整后EBITDA利润率约为20%。

储备管线扩张：合格数据中心项目储备管线从1.25亿美元增至超过2亿美元。管理层表示，大部分增量来自潜在新客户，其中包括数个价值约为5000万美元或更高的重大商业机会。

产能保障机制：MEC正与多家客户就专属产能承诺或付款担保进行积极磋商。公司还在就客户造成的延期、额外人员配置及营运资金需求寻求补偿。

工厂与用工策略：MEC正根据劳动力可用性和生产需求在各工厂之间调整项目布局。管理层重点指出，黑泽尔帕克（Hazel Park）、芝加哥、北卡罗来纳州和佐治亚州等地的招聘环境更为有利。

长期产能：基于已获项目和当前储备管线，管理层预计MEC将在18个月内达到满产状态。汤姆森工厂及计划中的设备投资旨在将总营收产能拓展至此前披露的约8.50亿美元水平之上。

财报电话会议完整实录

---

完整财报电话会议逐字稿

管理层陈述

Stefan Neely

Including the risks described in our periodic reports filed with the Securities and Exchange Commission. Except as required by law, we undertake no obligation to update our forward-looking statements. Further, this call will include the discussion of certain non-GAAP financial measures. Reconciliation of these measures to the closest GAAP financial measure is included in our quarterly earnings press release, which is available at metinc.com.

Following our prepared remarks, we will open the line for questions. With that, I would like to turn the call over to Jag.

Jagadeesh Reddy

Thank you, Stefan, and good morning, everyone. Our third quarter results were in line with our expectations, reflecting strong demand and accelerating project volumes within our data center and critical power end market.

Net sales increased approximately 18% year-over-year. This was supported by the continued growth in data center and critical power, the ongoing recovery in commercial vehicle and sustained demand in construction and access. In data center and critical power, additional programs reached full production during the quarter, improving capacity utilization and driving a sequential improvement in adjusted EBITDA margin of approximately 190 basis points.

At the same time, we remain in the early stages of ramping the more than $145 million in projects secured in this end market over the past year. The pace and scale of customer requirements continue to place significant demands on our people, equipment and organization.

On October 1, we closed on the acquisition of a manufacturing facility in Thomson, Georgia, executing the capacity strategy we outlined following our second quarter equity offering. I will discuss the facility and our plans with it in more detail shortly.

Operational performance across our plants remained uneven during the quarter for two distinct reasons. First, at the facilities supporting data center and critical power, customer program timelines continued to evolve rapidly, requiring our teams to adjust production schedules and resources in response.

We also incurred approximately $3 million of data center and critical power project launch and outsourcing costs, along with premium labor costs supporting other customer programs. Second, facilities more heavily weighted towards agriculture and powersports are operating below efficient utilization levels as customer volumes in those end markets remain soft. We are actively shifting programs across our footprint to better balance workloads and align our cost structure. This rebalancing takes time and the resulting underutilization weighed on margins in the quarter. These costs were in line with our expectations, though.

As discussed on our last call, we continue to view them as temporary. We expect them to moderate gradually as demand continues to accelerate and new awards continue to enter the launch cycle. We expect these ramp-related costs to decline through early 2027 as equipment on order is delivered, outsourced production returns in-house, and we continue to fill open positions. Importantly, we are absorbing these near-term costs deliberately because the programs we are launching today are accretive to our long-term margin profile.

Finally, the demand we see today reflects a multiyear investment cycle, and we believe MEC remains in the early stages of that opportunity. Our products are typically installed in the later stages of a data center build. This means the programs we are producing today support infrastructure investments that broke ground roughly three years ago. The level of investment already committed across the industry, combined with our less than 5% penetration among our OEM customers gives us confidence in the long-term demand outlook.

While near-term sentiment around the data center construction may fluctuate, we remain confident in the durability of this demand and the multiyear growth opportunity ahead of MEC.

Moving on to some of our key end markets. Commercial vehicle net sales increased nearly 35% year-over-year, reflecting improved customer demand, supported by an increase in North American Class 8 production. ACT continues to project Class 8 production growth of over 8% in 2026. While its outlook for 2027 has moderated, ACT still expects production to increase approximately 9% in 2027. We remain encouraged by current build rates and the activity we are seeing from our customers.

In construction and access, net sales increased approximately 9% year-over-year on continued improvement in nonresidential construction demand. We expect this end market to remain a steady contributor as nonresidential activity continues to improve. In powersports, net sales decreased by approximately 27% year-over-year, driven by continued soft customer demand as OEMs in this end market realigned their global supply chains. This trend is consistent with our expectations and is likely to persist into 2027.

Military remains constrained by ongoing program transition delays and agriculture remains in a cyclical trough with no near-term inflection expected. Both are consistent with the outlook we have shared previously. Within data center and critical power, net sales increased approximately 65%, reflecting accelerating demand from current customers and new project launches. This end market represented approximately 22% of total net sales in the quarter.

During the quarter, we secured approximately $45 million of new data center and critical power project awards, while our qualified opportunity pipeline now exceeds $200 million. With growing demand, we continue to evaluate opportunities for customers to reserve dedicated capacity as well as arrangements where customers fund certain capital investments required to support their programs.

We will continue to manage capacity and production priorities with a focus on sustainable, diversified and profitable long-term growth. These awards secured during the quarter reflect a combination of new customer wins and continued expansion with existing major customers. These programs span a range of power gen enclosures, switchgear, switchboard enclosures and uninterrupted power supplies.

We continue to expand our share with key commercial vehicle customers as they prepare for upcoming product launches tied to the 2027 EPA regulatory changes. These programs are expected to begin entering production in late 2026 and continue ramping into 2027. Beyond commercial vehicle, we secured new business tied to model introductions with a powersports customer while also expanding our business with an agricultural customer.

Turning to capital allocation. Our focus remains on three priorities: investing in organic growth, continuing to reduce leverage and pursuing selective accretive acquisitions. First, organic growth. Based on our awarded programs and pipeline, we currently expect to reach full capacity within the next 18 months. We are deploying capital now to expand our footprint ahead of that point.

On October 1, we closed on the acquisition of our Thomson, Georgia facility located in the Southeastern United States. Closer to key data center and critical power customers, this facility establishes a platform for growth in a region where our capacity has been constrained. Together with planned equipment investment, we expect this facility to add approximately $50 million of annual revenue capacity by 2028. This will extend our total revenue capacity beyond the approximately $850 million we previously disclosed. Rachele will cover the associated capital spending and the return criteria we apply to these investments.

Second, deleveraging. We ended the quarter with a net leverage of 2.9x and our long-term target remains 2.5x. As volumes ramp and profitability improves, we expect earnings growth and cash generation to drive continued progress toward that target. Third, accretive M&A. We will remain opportunistic in pursuing acquisitions that strengthen our competitive position, expand capacity and support long-term value creation. Looking ahead, we are confident in the opportunity in front of us and in our ability to deliver profitable growth in 2027 and beyond.

Before I turn the call over to Rachele, I want to address the trajectory of our margins as we navigate this period of transition for MEC. In the past, we have pointed to an approximately 15% adjusted EBITDA margin as the target for our business. We continue to see a clear path to that level and potentially beyond. That path rests on four drivers, each of which are already in motion.

First, the moderation of ramp-related costs as equipment on order is delivered and placed into service and outsourced work returns in-house. Second, improved absorption and operating leverage as our data center and critical power programs continue to ramp toward full production. Given the scale of demand in front of us, we are intentionally adding people, equipment and capacity ahead of our current levels. As volumes catch up to that resource base, we expect operating leverage to significantly improve.

Third, the continued build-out of our workforce. Labor availability remains tight in certain facilities. As hiring catches up, we expect to reduce our reliance on premium labor and overtime. Lastly, the normalization of demand in our agriculture and powersports end markets. Lower customer volumes in these markets are pressuring margins through underutilization, as I noted earlier. The footprint rebalancing already underway should help offset that and the recovery in those volumes would provide additional margin support.

These 4 drivers, combined with the continued mix shift toward higher-value data center and critical power programs as this end market moves towards 25% of total revenue gives us confidence in the trajectory. We expect to make measurable margin improvement over the next 12 to 18 months, and I want to be clear about the shape of that progress.

These drivers work on different timelines. Ramp-related costs should moderate first beginning in early 2027. Hiring and absorption improve as programs reach run rate over the course of the year and the timing of recovery in agriculture and powersports will be determined by demand in those end markets. As a result, we expect 2027 to be a year of continued growth with margins improving sequentially rather than in a single step. The benefit of this work will be most visible as we exit the year and move into 2028.

We will provide our formal outlook for 2027 on our fourth quarter call. We believe these actions position MEC to deliver attractive, sustainable returns on invested capital as we capitalize on a multiyear secular growth opportunity.

With that, I would like to turn the call over to Rachele.

Rachele Lehr

Thank you, Jag, and good morning, everyone. Total sales for the third quarter increased 18.1% on a year-over-year basis to $170.4 million. Manufacturing margin was $18.8 million or 11% of net sales compared to $15.9 million or 11% of net sales in the prior year period. The year-over-year increase in manufacturing margin dollars was primarily attributable to improved capacity utilization driven by higher sales volumes in data center and critical power as well as growth within commercial vehicle and construction and access end markets. This increase was partially offset by $3 million of data center and critical power launch and outsourcing costs and premium labor.

Other selling, general and administrative expenses were $8.8 million compared to $10.5 million in the prior year period. The decrease primarily reflects a decrease in nonrecurring acquisition-related expenses and the realization of synergies associated with the Accu-Fab acquisition.

Interest expense was $3 million compared to $3.4 million in the prior year period due to decreased average borrowings under our revolving credit facility, partially offset by higher interest rates in the current period. Adjusted EBITDA was $17 million or 10% of net sales compared to $14.1 million or 9.8% of net sales in the prior year period. The increase was primarily due to higher sales volume and improved capacity utilization, partially offset by data center and critical power project launch and outsourcing costs and premium labor. On a sequential basis, adjusted EBITDA margin improved by approximately 190 basis points.

Turning now to our cash flow and the balance sheet. Free cash flow during the third quarter of 2026 was a use of $3.3 million as compared to a use of $1.1 million in the prior year period. The decrease was primarily driven by higher capital expenditures related to equipment needed to support rapidly accelerating demand in the data center and critical power end market, partially offset by higher cash flow from operations. As of September 30, our net debt was $146.6 million. The ratio of net debt to trailing 12-month adjusted EBITDA calculated per our current credit agreement was 2.9x.

Now turning to a review of our updated outlook for the full year 2026. For the full year of 2026, we currently expect net sales of between $630 million and $640 million, adjusted EBITDA of between $48 million and $52 million and free cash flow use of between $6 million and $16 million. This outlook assumes a full year of Accu-Fab ownership, incremental cross-selling revenue and continued recovery in commercial vehicle and construction and access, partially offset by ongoing softness in our agriculture and powersports end markets. Our guidance also incorporates the impact of continued launch and outsourcing costs associated with data center and critical power programs, along with premium labor.

Our updated outlook reflects two factors. First, as Jag noted, customer program timelines in data center and critical power continue to evolve and the timing of certain project volumes has shifted into early 2027. Second, mix within our legacy end markets is weighing on margins. Strong growth in commercial vehicle is occurring alongside continued softness in agriculture and powersports, where lower volumes are leaving portions of our footprint underutilized.

As a result, we expect fourth quarter adjusted EBITDA margin to be modestly below the third quarter level. We view this as a transitional quarter, and it does not change the path Jag outlined with ramp-related costs beginning to moderate in early 2027 and margins improving sequentially from there. Free cash flow guidance reflects working capital investment and planned capital expenditures of $25 million to $35 million, including the initial investment in the Thomson, Georgia facility.

As Jag mentioned, on October 1, we completed the acquisition of our Thomson, Georgia facility. Inclusive of the purchase price, facility improvements and production equipment, we currently expect to invest a total of approximately $25 million in the location through 2028. Of that amount, we currently anticipate approximately $10 million of capital expenditures in 2027, with the remaining to be invested in 2028. When fully deployed, the investment is expected to support approximately $50 million of annual revenue capacity.

The timing of the spending will be aligned with equipment lead times, customer demand and the ramp of awarded programs. We will continue to evaluate each phase of the investment against our established return criteria, including an internal rate of return of at least 15%. This phased approach allows us to add capacity as demand develops while maintaining discipline around capital deployment and cash flow.

In summary, our third quarter reflected the dynamics Jag described, strong volume growth and improving utilization in data center and critical power, partially offset by launch costs and softness in a couple of our legacy end markets. Looking ahead, our focus remains on converting higher volumes into improved profitability and cash flow. While we are not providing 2027 guidance today, we expect that improvement to build gradually over the course of the year. Ramp-related costs should moderate first with the full benefit of improved fixed cost absorption, operating leverage and footprint rebalancing becoming more reflected in our results by late 2027 or early 2028.

We will pursue that progress while maintaining a disciplined approach to working capital and capital spending. We remain focused on reducing leverage toward our long-term target of 2.5x, while funding organic growth opportunities that meet our return criteria. We will also remain selective in evaluating additional acquisition opportunities. With that, operator, we are ready to open the line for questions.

Operator

[Operator Instructions] Your first question from the line of Mike Shlisky with D.A. Davidson.

分析师问答

Michael Shlisky

So wanted to just get a little more detail on the margin improvement that you discussed that will be taking place throughout 2027. I guess I'm trying to figure out what is the base or the bottom of the margin progression. Is that what we just saw? Although I guess it was up quarter-over-quarter, but is it what you expect to see with fourth quarter as being the base where you'll go up, up, up? Will there be a rough first quarter, second quarter history that comes from there up and up in the third and fourth quarter? Just want to see the exact trajectory or the ups and downs, and where the absolute lowest point of this is.

Rachele Lehr

Mike, yes, so I think you're correct. Q4, as you can see, is going to be down slightly from Q3. That reflects seasonal softness and some of the shifting of the customer timelines, and then we still have the DCP ramp costs.

As we move into Q1, we expect to be able to have the capital that we purchased about six months ago up and running, and that will allow us then to reduce our outsourcing. And then that will also allow us to start reducing our premium labor. So that will happen sequentially in the first part of the year, such that in the second part of the year, we will have the equipment up and running. We will have the labor to support it, and then we'll sequentially improve as data center ramps throughout the rest of the year.

Michael Shlisky

Okay. So it's 4Q, that's kind of the base. That was the main question.

Rachele Lehr

Yes.

Michael Shlisky

Yes. Then secondly, I mean, it's a very fresh news, but wanted to get your comments on what happened in elections last night. I guess it's a two-part question there. First, any changes within the state of Wisconsin or your other major tax bases? Any changes in party or within those states and those jurisdictions that you're looking at or thinking about?

And then more broadly, there's some of the candidates for various positions in Congress and governor offices, they made a lot of comments about data centers and critical power as not being what they wanted to see within their states, and things along those lines, wanting to cancel or revoke projects or slow things down, et cetera.

Just curious whether the projects that you've been working with, companies that you're working with, do they all still feel fairly confident about how the data center and critical power build-out will take place, barring the very current near-term timeline changes? But more broadly, do folks still feel confident when you think about the broader build-out here?

Jagadeesh Reddy

Yes. Great question, Mike. I've been thinking about that since yesterday. Going into the election, right, there was significant opposition to data center build-out in multiple states, whether red or blue states didn't matter, right? There was a groundswell of opposition to data center build-out.

But if we all step back and think about it, what was that about, right? It was really about many of the state governments providing tax breaks to the data center operators, but also how the electrical infrastructure investments were being financed, who was paying for it and then how the electricity costs are being shared, whether the local customers were going to bear the brunt of that cost increases, right? So that was the main reason why many of the states started putting moratoriums, et cetera.

But again, I'm not a futurist, but sitting here, taking it all in, right, my -- at least the base case is that long term, we need data centers to continue to be built out. No one disagrees with that. It's the matter of what's going to be the federal policy, what is going to be the state reactions to sharing of infrastructure investment costs and power generation and also the actual sharing of the cost of electricity that is being provided to these data centers.

And of course, there's a bunch of misinformation about water usage. Many of these data centers, right, actually recycle their water. We're sitting here on Lake Michigan, right? It's not an issue at all in some of the states. But I think data center operators will get over or at least provide the right information on water usage, right?

So if you take that all in, I am very, very bullish on continued investments in data center build-out for the foreseeable future. Internally, we were joking about that, hey, we said in our prepared remarks, we could run out of capacity in the next 18 months and someone said, maybe we'll run out of capacity in 19 months instead of 18 months, right? So I think that's our current position. End of the day, given the tightness in the debt markets and some of the rationale or rather return on investments for the data center operators and the hyperscalers, that is going to drive whether the pace of build-out is going to be rather than we're going to need to build data centers or not.

So again, I could not be more bullish on data center build-out. This election results changes nothing for us. And in fact, right, I'm actually cautiously optimistic that post-election, whether the states or the federal government, right, will come back to the table with some guidelines, some regulations and streamline approvals of the data centers, streamline tax abatements or whatever else that they want to provide in the communities and put some guardrails around right, who's going to pay for infrastructure investments and power usage and cost of electricity to these operators, right?

Having said all of that, if you really step back and then say, outside of building the actual building and the chips, right, where is the rest of the investment is going to be? And it's going to be in the electrical grid, electrical infrastructure, electrical equipment, switchgear, that's exactly where we're playing, right?

All of the product lines that in DCP that we're supporting and growing with all relate to electrical infrastructure, right? So I feel really great about how we have positioned MEC to take advantage of this build-out that's coming for the next three to five years.

Michael Shlisky

Great. The other part of my question was about what is going on in maybe Wisconsin? Are there any candidates, one, that were demanding for higher corporate tax rates, stuff like that? Or any other states where you have a very.

Jagadeesh Reddy

Yes. So look, I will add another one, right? We don't show up until year three of the build-out. So many of the programs that we're currently supporting those projects broke ground three-plus years ago. So we're not concerned about, right, what might happen in the next year or so, things will settle down.

Again, I don't want to get into politics, but I think the governor elected in Wisconsin is actually very moderate in terms of his view on data centers. But also, right, we're not just supplying into Wisconsin data centers. We're supplying across the country. So as I said earlier, right, we expect this noise to come down a bit and then more cooler heads prevail because we do need these investments. We're also going to put these data centers in China, right? That's -- so we need them, we're going to get them, and then there will be a lot more streamlined process going forward.

Michael Shlisky

I apologize, Jag. I was actually just asking about broader corporate tax rates on your entire income within Wisconsin, and what that might mean -- if anything.

Rachele Lehr

We don't anticipate really any significant changes. Wisconsin was blue. It's remaining blue. So as far as that goes, we expect that really to continue to be the same.

Operator

Your next question from the line of Andy Kaplowitz with Citi.

Andrew Kaplowitz

Jag, as you know, your exposure to data centers is continuing to get larger. So how, if at all, can you better manage the variability of data center starts as you go into '27? And so maybe you can improve efficiency. And is there a way to aggregate how much of the impact on your '26 and Q4 EBITDA pressure is from this timing of efficiency versus maybe the other issue you talked about, which is ag and powersports underutilization?

Jagadeesh Reddy

Yes. I'm going to let Racheal jump in, but I'll just share that. That's a really good question, Andy. That as new entrants into this market, right, we're still learning on how the schedules continue to evolve for some of our DCP customers. I would say that about $20 million approximately of DCP revenues pushed out of second half of '26 into 2027.

So I mean the reason -- I would say, the main reason why, in addition to just seasonality, the main reason why we came down on our Q4 margins is really because of this highly profitable data center revenues pushed out into first half of next year because our customers are continued -- this is more than one customer, are continuing to deal with their own supply chains and their challenges on quickly ramping up to these high volumes. So with that, let me turn it over to Rachele.

Rachele Lehr

Yes. I think Jag just hit on the biggest piece of it. If you take $20 million at 20% adjusted EBITDA, that's a significant piece of that change. You throw in a little incremental labor and launch costs as we have not gotten to the point of full run rate production on that, that really brings our midpoint down from that $56 million to that $50 million. So those factors really are what contribute there. as we continue to move forward and really hit that DCP ramp, as we just talked about, we'll see sequential improvement as we move through 2027.

Andrew Kaplowitz

Very helpful. And then, Jag, you mentioned now over a $200 million pipeline of opportunity for data centers. I think you previously said data centers could eventually be 25% to 30% of your mix, it could actually be higher than that. And if you think about sort of protecting that from postponements, I know you're going to say you're three years after these things start. So hopefully, they don't get postponed. But this whole idea of locking in capacity with certain customers, are you closer to that? Do you think you could sort of lock in a lot of your incremental capacity over the next 12 to 18 months?

Jagadeesh Reddy

I continue to be optimistic on locking our capacity with our customers, Andy. We have more than one customer that we are in active conversations with to lock capacity or have some sort of capacity payment guarantees. So we're cautiously optimistic that, that will happen.

At the same time, the pipeline continues to grow. I am more concerned about when do we have to raise our hand and then say we can't take any more new business because as you can imagine, right, these programs, unlike our industrial programs, which could take 18 to 24 months to show up as revenue, we can start producing production parts in three months. And within six months of award, right, we're in a full ramp for many of the DCP customers.

So that means at least for the first three to six months, I have to protect that capacity once we win a program because I know that the ramp is coming in that particular client for that particular customer, right? So we're working through all of those mechanics and then there's obviously cost to hold that capacity, not only for that six months before the full ramp hits, but also any further program delays.

So as we're learning how to manage that better, the sales teams are getting smarter, getting more aggressive. So we're going back to DCP customers, I would say, even in Q4. And not just talking to them, but actually charging them for delays, charging them for incremental headcount, charging them for working capital. So we're getting better at this.

As we go into 2027, I feel really good about how we will be able to manage, how we manage our costs. Look, we've been in business for 81 years in the industrial markets. We've been in DCP for a year, and it's a highly fast-moving industry. So there were definitely some learnings in 2026 that will apply going into 2027. I could not be more bullish about 2027.

With all the things that the teams have done, what we have put in, the incremental machine, capacity that's coming online in Q1 and where we are in our hiring cycle, I feel very good about protecting our capacity, being able to charge for our capacity and potentially locking our capacity with a couple of customers.

Operator

Your next question from the line of Ross Sparenblek with William Blair.

Robert Samuel Karlov

This is Sam Karlov on for Ross. I just want to dig into the qualified pipeline for a second. I mean it jumped up from $125 million last quarter to $200 million. Can you give us some color on why this jumped so sharply? And what kind of opportunities kind of got added? Were there large opportunities, new customers? Or is it kind of more opportunities from existing customers?

Jagadeesh Reddy

I would say that most of the jump is from new customers that we have not done business with in the past. So we -- obviously, we won't talk about specific customers on these calls, Sam. But there's one customer that's got more than $50 million worth of three different programs. We have another European switchgear customer that's got another $50 million approximately of new programs. And we have another U.S. customer over $60 million worth of multiple programs, right?

So I would say that over the last three months or so, we were able to continue to work with these customers, better qualify. And many of these programs that these customers I just mentioned are very close to decision, right? So now we can say that not only the customer is very close to decision, our chance of winning those programs is greater than 50%. And that's when we feel really good about putting that and actually raising the bar on our qualified pipeline.

Robert Samuel Karlov

Got it. And just a follow-up on that. Can you talk about what kind of the primary decisions or the factors that are leading to how these customers make decisions on where they kind of put this volume?

Jagadeesh Reddy

Yes. So two things. One, location, location, location. So that is getting to be even more important and hence, plays well with our Thomson investment. Second one is the ability to bring on a $50 million program in a three to six month period, there are not a whole lot of fabricators in North America that could do that and do that effectively. So that really plays to our strength because not only we have the footprint, we've been investing. Yes, we're paying the cost for it upfront. You've seen how we have outsourced, how we have hired temp labor to bring in those capabilities quickly. We could have waited until 2027 to invest in these resources and capacity, and we would have missed the boat on some of these opportunities, right?

So -- even though in the short term, right, it feels like there is a drag on the P&L, which it is, I feel really good about what this is going to do to us, not only in 2027, but also 2028 as we win these significantly larger programs. When we started after the acquisition of Accu-Fab, we were winning $2 million, $3 million, $5 million programs. Now we're competing for $20 million, $30 million, $40 million, $50 million programs, right? That's the difference.

And there are not a whole lot of folks on the Fab 40 list that can step up and then say, not just one or three programs, but we're competing easily 7 to 10 different programs that range from $20 million to $50 million in size, right? So that's a testament to MEC's experience, capabilities, footprint and our investments that we have made. I want to thank all of our investors for standing by us in our equity raise. We're putting that equity raise funds to good use in investments for the future.

Robert Samuel Karlov

Got it. No, that's good to hear. Just a quick one on CV. I mean, I noticed you guys put out your kind of preliminary 2027 expectations. Commercial vehicle expecting up mid to high single digits. I mean, comparing that with the ACT growth expectations for 9% next year, can you maybe help us think about how your kind of content per truck moves up with the 2027 model year and kind of how that should flow through to revenue?

Jagadeesh Reddy

Yes. You can also see third bullet, right? We anticipate some strategic -- exits in the CV market. As we talked about in the past, in the industrial markets in general, I'll be specific to CV and ag and also construction and access, we are aggressively pursuing higher pricing from many of our customers. We expect net-net $20 million to $30 million worth of exits next year from our industrial customers that just gives us the opportunity to redirect that capacity towards data centers.

So even though ACT is projecting 9% growth next year, let's say, close to 300,000, approximately 300,000 Class 8 builds next year, we agree with that. At the same time, right, we might see some exits that will help us in the long run to reposition our capacity towards higher-margin data center work.

Operator

Your next question from the line of Greg Palm with Craig-Hallum.

Greg Palm

Maybe to start with, I was hoping you could unpack a little bit more about some of the, call it, the inefficiencies, the ramp costs, some of the schedule changes, all that stuff that's kind of impacting timing. And as we think about labor or lack thereof, some of the availability challenges, does that impact where you think you want to put a lot of this data center work relative to what you were thinking six or none months ago?

Jagadeesh Reddy

Yes. I'm going to start and then Rachele will jump in. That's an excellent observation, Greg. Where we thought we would have some of these programs six, nine months ago is different than where we're putting those programs in today. Many of our rural plants continue to struggle to ramp up. Across the company today, we have somewhere between 230 to 250 job openings in our plants. That number continues to go up because we continue to win new business. That's not a challenge, obviously, as well.

But even if we're static, we still have 200-plus job openings in the network. So what that did was to support in the short term, we went the route of outsourcing, then we started hiring temp labor in many of these locations. As everyone knows, welders are the hardest trade to hire right now. So we have a couple of plants where we are today paying $90 an hour plus per diem, whereas the starting wage in our factory today at the same factory is around a little below $30, right?

So our customers across the street at the same plant are also paying $90 an hour. So it's a known problem, right? So what are we doing about it? Number one, we restarted in multiple plants, boot camps to train our own welders so that we can promote within, train folks that are maybe doing a different job that can make -- maybe they were making $18 an hour, now they can make $30 an hour, right?

So we're doing some of that work. But more importantly, in that particular plant, right, we decided to pull out some of the work that had weld content because we cannot hire welders at that location. And we started putting in actually data center work that does not have welding, right? Many of our data center projects don't have a lot of welding. So we're rebalancing the network where, where is the labor available? Let's put labor -- sorry, programs that can be tailored to the labor availability and then move programs around, right? We did a lot of that work in the second half -- sorry, in Q3, in particular, and that work will continue through Q4. Again, that's what is going to help us going into 2027.

At the same time, we know that we can hire a lot of people in Hazel Park, in Chicago, in North Carolina and now in Georgia, right? So we're going to focus heavily towards those four locations where we can hire many folks in all kinds of trades at these locations. So we're shifting our focus towards those locations and reordering where do we put weld content, where do we put assembly type content, where do we put different type of programs, right?

So all of that work is underway. The team has done a phenomenal job in Q3, and we feel really good about entering 2027 in mitigating some of these constraints, right? So the machine constraints are going to come down by Q1 with new machines now being installed in many of our plants that we paid for almost two quarters ago. And now some of the labor constraints will continue to come down as we go into first half of next year, exiting Q2.

And last but not least, the outsourcing will come back in as the machines come online, right? So that's how we're managing some of these challenges. But as you said, right, yes, that changed what our outlook was six, nine months ago and where these programs are going to go to where we're slotting in these new programs and new wins across the network.

Greg Palm

Yes. Okay. That makes sense. And then kind of related to that, your ability to reprice both kind of existing contracts and certainly some of the newer stuff in light of whether it's just higher inflationary costs, higher labor costs, et cetera? Like what are you seeing?

Jagadeesh Reddy

Yes. We are taking a strong position on repricing. As I answered Sam's question earlier, we do expect some strategic exits, mutually agreeable exits, if you will, and we're going to leverage price to do that because we do need some incremental capacity in some of these plants, and we're using price to exit some of our legacy programs and perhaps put in some new programs.

Operator

[Operator Instructions] Your next question from the line of Ted Jackson with Northland Securities.

Edward Jackson

I was a little concerned that they didn't recognize that my hand was up. Most of my questions are answered. I have a couple. I wanted to talk a little bit about ag and powersports. And I mean, powersports is a little bit of its own beast. My sense with regards to ag, though, I mean, you look at it, Deere has been hiring people and talking about kind of bringing production up in components of ag. If you listen to AGCO's call, North America was somewhat of a bright spot. Most of these guys have been producing below retail demand in North America and expect to not necessarily see retail grow, at least not that much, but they do expect to see volume improvements simply because they'll be able to produce to demand instead of beneath it. Yet you're showing AGCO go down.

And then over on powersports, if you kind of dance through powersports, it seems like some of the motorcycle side of things in the U.S. is kind of stabilized, performing well. The side-by-side market, maybe it's not doing fabulously well, but if you kind of parse it through, and maybe I'm wrong on this, but take out the recreational side and look at more the utility side, and I assume that all kind of runs through that same line item for you, that's definitely a bright spot. I guess where I'm going with this is are you --is the declines that you're putting forward with regards to your outlook for those markets, am I missing something, or is it more because you're walking away from business that [indiscernible]? It doesn't square -- the circle doesn't square very well.

Jagadeesh Reddy

Yes. Good question, Ed. Let me address one at a time. Let me address ag first. We have to see -- from a macro perspective, we have to see tariff relief, and we have to see diesel prices come down for the economics to work for a farmer to afford a brand-new large HP tractor or a new combine, right? So that is the challenge that I don't see that happening. I don't know, maybe last night's results might change some of that next year. Maybe we're cautiously pessimistic about it, and we'll see how it plays out next year.

But some of the hiring you talked about at one of our customers, in fact, our mix of business with that customer predominantly tilts towards construction rather than ag. So ag continues to be a smaller and smaller portion of our overall sales. So we're not putting a lot of weight on an ag recovery in the short run. But whenever it returns, right, we will be prepared for it.

In terms of powersports, you're absolutely right. We're sitting here in Milwaukee. This morning's news talked about significant sales in our local customer. At the same time, that same customer in-sourced eight jobs to the U.S. and that pat on their back. And what that really means is that they're outsourcing more and more of the work similar to every other powersports customer to Asia. We're seeing significant movement in the supply chains where they're relocating to Asia.

So given that, even though the market may be stabilizing in terms of retail sales, I think from a supplier perspective, right, we continue to see declines in our business with these powersports customers. And hence, our focus -- we had two powersports-specific factories, we already converted one to a data center-focused factory. And then we continue to look for opportunities to put in more data center work into the other powersports centered factory as the plan progresses here.

So I'm not very bullish on powersports returning, at least for us, and we'll see what happens with ag.

Edward Jackson

Okay. That's a great answer. Shifting over on kind of some of the outsourcing stuff. You commented in the release that you had a headwind of about $3 million in the third quarter. And if my memory is right, that was kind of what you had expected for all of the second half of '26. So it seems like that was a little bit more maybe in the quarter than was expected. And so I guess where I'm going with that is, am I correct on that and then that headwind that you're having with some of the costs as it relates to outsourcing of products, what do we expect for fourth quarter?

And I mean, I know you're not giving '27 guidance, but maybe kind of how we should think about that headwind in the first quarter as well. That's my next question.

Rachele Lehr

Yes. The -- so as we shared coming into this, we expected $1 million to -- that's really launch and outsourcing. So you almost have to look at it, the two added together for what we reported. We had previously shared that those would be each about half of that amount. And so that's where the disconnect is. So we haven't changed our total amount on outsourcing. We had projected second half for that would be $2 million to $3 million. So you're absolutely right. It's about half of that in each of the quarters. It's the aggregate of the two, that's the $3 million.

Edward Jackson

Okay. So that's my confusion is --it's an apples and oranges thing in terms of what I was looking at.

Rachele Lehr

Yes.

Edward Jackson

And then my next question is just on -- you've ordered a bunch of equipment. It's all being delivered. You're having -- you're installing it. So I guess the first question on this is walk us through the timing of that. Do you expect to have this equipment installed by the end of this fiscal year or installed by sometime in the first quarter?

And then how long will it take you to kind of get it installed, commissioned and start -- being able to produce off of it because that brings down two things for you. And one of them is just this outsourcing revenue, and then two is better asset utilization rates. So a discussion with regard to kind of the timeline for that equipment. And then maybe a discussion with regards to how you're thinking about CapEx for '27.

Rachele Lehr

The equipment is -- some of it has arrived or is arriving late in Q4 to really have it up and running, that's going to be sometime in Q1. So to your point, once we get that up, that will solve a few of our problems. It will take time, though, to transition the outsourcing work back in once it's up and commissioned, but really see Q1 as being a key player in having the equipment up and being able to move through those costs.

And then as we look to 2027, some of these -- this equipment will obviously have to make a final -- as it arrives, final payments and final pieces. But as we've shared on our presentation and our guide, we're looking at $40 million to $50 million next year, still focused on the growth. That's continued to be the big area that we are focusing on as we continue to grow in the data center and critical power space.

Edward Jackson

Okay. So when we think about the equipment that you have, you're going to have it all in place in the fourth quarter. You'll get it commissioned in the first quarter, but we won't really see the impact from that progress in margins probably until the second quarter of '27-- and they are part of the scale go forward. And then -- so that means that the first quarter, what you're really expecting to drive your margin higher is this $20 million of revenue that slipped out of the fourth quarter?

Rachele Lehr

Yes, that's our -- improving the run rate on the DCP and getting that as a higher piece of our mix will certainly help our overall margins.

Operator

This concludes our Q&A. I will now turn the call back to Jag Reddy for closing remarks.

Jagadeesh Reddy

Before we close, I want to thank our employees for their dedication and hard work this quarter and our shareholders for their continued trust in MEC. We are confident in the path we're on and in our ability to build a stronger, faster-growing and more profitable company in the years ahead. We look forward to updating you on our progress. Thank you for joining us today.

Operator

This concludes today's call. Thank you for attending.