DVMT Tracking Stock: Silver Lake's Creative Financing of the $67B EMC Deal

Tracking stocks were considered a 1990s relic by the time Silver Lake and Michael Dell got to the EMC negotiating table in 2015. The last famous one — AT&T's tracker for its wireless business — had been collapsed in 2004. Investors didn't like them, courts didn't love them, and banks had stopped underwriting against them.

Then Silver Lake and Dell quietly issued $20 billion of fresh tracking stock as part of the consideration for EMC. It was the largest tracking-stock issuance in the security's history, and it was the only reason the $67 billion deal — by far the biggest tech LBO ever attempted — was financeable at all.

Here's how the structure actually worked, why no conventional financing could have replaced it, and how Silver Lake collapsed the whole thing two years later in a reverse-merger exit that returned roughly 2.0× their tranche.

What a tracking stock actually is

A tracking stock is a class of equity issued by a parent company that is designed to track the economic performance of one specific subsidiary or business unit — without legally separating that unit from the parent.

The mechanics:

• The parent issues a separate class of common stock (often a "Class V" or similar designation).
• The dividend rights, and theoretically the residual claims at liquidation, are pegged to the cash flows or value of the tracked business unit.
• The tracked business itself does NOT spin out. It stays inside the parent, consolidated on the parent's balance sheet, with the parent retaining 100% legal ownership of the assets.
• Voting rights on the tracker are usually limited — holders track the economics but don't elect the parent's directors.

Think of it as a synthetic spinoff: the parent gives you a claim on a piece of the business without actually carving that piece out. It's useful exactly when you want the market to value a subsidiary at a premium to where the parent trades, but you don't want to take the tax hit, lose operating control, or trigger debt covenants by actually separating it.

That's the textbook definition. The Dell-EMC version is the only place at PE scale where the structure has been used to bridge a megadeal that couldn't otherwise be financed.

Why the $67B EMC deal needed one

EMC's most valuable asset by far was its ~81% ownership stake in VMware (VMW), which traded as a separately-listed public stock. When Dell and Silver Lake sat down to construct the bid, the cap-table problem was severe:

If Silver Lake had simply tried to raise $67B in cash plus conventional LBO debt, three things would have broken:

1. The cash equity required would have been impossible. With $24B of sponsor equity already at the absolute edge of what Silver Lake plus Dell plus MSD Partners could write, finding another $20B in real cash to cover the VMware portion of EMC's value would have required outside co-investors no one wanted in the syndicate.

2. Selling the VMware stake to fund the deal would have crushed tax efficiency. EMC's basis in VMware was very low. Disposing of the stake to monetize it for purchase-price funding would have triggered a corporate-level capital gain on the order of tens of billions before any of the proceeds reached the deal.

3. Buying VMware outright would have doubled the deal size. VMware's public-market cap at the time was its own megadeal. Folding it into the take-private would have pushed total consideration past $90B — financing that quite literally didn't exist as a single underwriting in 2015.

The DVMT tracking stock solved all three problems simultaneously. Dell handed EMC shareholders a paper claim on VMware's economics in lieu of $20B of cash, kept VMware itself trading independently inside the new Dell Technologies structure, avoided the disposition tax, and brought the cash equity requirement down to a level Silver Lake plus Dell could actually fund.

The numbers at close

Per Dell's 13E-3 proxy filing and contemporaneous WSJ reporting, the closing capital structure looked like this:

A standard LBO would have shown sources as some combination of bank debt, sponsor equity, and (occasionally) seller rollover. The DVMT line item is the structural novelty: it sits on the sources side of the table as deal consideration, but it isn't cash and it isn't traditional debt. It's a security backed by the most valuable piece of the asset being acquired — issued by the acquirer, paid to the seller's shareholders, and traded publicly the day after close.

If you want to think about how that fits into a normal sources-and-uses framework, the same accounting discipline applies: every dollar of consideration on the uses side needs a corresponding source, even if that source is a non-cash claim on assets the buyer already owns post-close.

How DVMT actually traded

Here's where the elegance of the structure ran into market reality. DVMT was meant to track the value of Dell Technologies' VMware stake. In a frictionless world, the price of one DVMT share would have tracked the price of VMW (adjusted for the ownership ratio) very closely.

In the actual world, DVMT traded at a persistent and substantial discount to its theoretical NAV based on VMW.

A few reasons:

• No legal recourse to the underlying shares. A DVMT holder owned a claim issued by Dell Technologies. They did not own VMW. Dell's board retained discretion over how the tracked economics would be distributed, converted, or eventually unwound — and the market priced that discretion as a haircut.
• Limited float and concentrated capital structure. Dell Technologies was a private, sponsor-controlled company with multiple share classes. DVMT was the only publicly traded class with any meaningful liquidity, and its overhang was clear.
• Conversion ambiguity. The terms allowed Dell to redeem or convert DVMT under specific scenarios — at prices the market didn't trust would equal fair NAV.

The persistent discount created a real arbitrage trade. Hedge funds went long DVMT and short VMW, betting the discount would eventually compress when Dell collapsed the structure. The trade got crowded, lawyers got interested, and shareholder litigation followed alleging that Dell engineered the entire structure to extract value from DVMT holders during whatever eventual unwind the board chose.

That litigation overhang is part of the cost of using tracking stock at this scale. Conventional debt doesn't get sued by the people who hold it. Tracking stock — especially when the issuer controls the terms of its eventual collapse — does.

How Silver Lake actually exited

The exit was a reverse merger, not a conventional sale or IPO. Silver Lake's full return cycle on the EMC investment:

The 2018 reverse merger is the part of the structure most worth understanding. Rather than IPO Dell Technologies through a traditional underwritten offering, Dell offered DVMT holders a mix of cash and newly issued Dell Technologies Class C common stock in exchange for their DVMT shares. The deal was structured so that the same transaction (a) collapsed the tracking-stock overhang, (b) took Dell Technologies public, and (c) gave Silver Lake marketable public equity it could begin to monetize.

For Silver Lake's tranche, the realized return profile:

The 2.0× over 2.3 years math is a useful reminder that fund-tier IRR doesn't require a heroic MOIC. A short hold at a moderate multiple often beats a long hold at a high multiple when LPs measure you on annualized return. Silver Lake's reported 25% reflects the actual cash-on-cash timing across multiple tranches and the way realized returns get netted for LPs — but the headline insight holds. Quick exit, moderate MOIC, fund-tier IRR.

The Dell-side economics were different. Michael Dell held through the 2018 re-IPO and through the November 2021 VMware spin, capturing post-IPO appreciation that Silver Lake mostly didn't underwrite at entry. That's the structural difference between a financial sponsor's tranche and a strategic-plus-management tranche on the same deal.

What an LBO modeler learns from this

The DVMT structure is too specific to copy. Most associates will never model a tracking stock, and most deals don't have a publicly traded sub-asset of the target that's bigger than the parent's standalone value. But the principles generalize:

• When cash equity is the binding constraint, structural creativity beats raising more capital. Silver Lake didn't try to syndicate another $20B of co-invest. They engineered a security that substituted for cash equity using assets the buyer would own post-close.

• A "non-cash source" on the S&U is still a real source. Tracking stock, seller rollover, earnouts, contingent consideration — all of them appear in real deal structures and all of them have to balance the uses side of the table. The principle is the same whether you're modeling a $500M software buyout or a $67B megadeal: the cash equation has to close.

• Quick exits via reverse merger or re-IPO can outperform long holds. Silver Lake exited in 2.3 years at 2.0×, generating a 25% IRR. Most PE funds underwrite to 5-year holds at 2.5× MOIC. The faster, smaller deal cleared the fund-IRR hurdle by a wider margin than most slow-and-steady plays. The reverse merger collapsed an unusual structure quickly, which is part of why the IRR shape worked out.

• Litigation risk is a real cost of unusual structures. Tracking stock, dual-class voting, complex conversion terms — they all create legal exposure that conventional capital structures don't. That cost belongs in the underwriting case as a real line item, not just a footnote.

If you want to pressure-test how entry multiple, leverage, and hold combine to produce IRR on a deal of any size, the free LBO returns calculator runs the debt-paydown loop and surfaces MOIC and IRR live. Drop in 6.1× entry, 5.5× leverage, a 2-year hold, and a 7× exit multiple and you'll see roughly the shape of what Silver Lake underwrote on the EMC tranche.

For a deeper diligence build — including a multi-tranche capital structure where you'd actually model cash debt, rollover equity, and non-cash consideration in parallel — the All-in-One PE Model handles the structure end to end. The S&U builds the cap stack from seven tranche types, the debt schedule amortizes financing fees correctly, and the returns waterfall accounts for the difference between cash equity at close and equity at exit. You can see the LBO sheet preview before buying.

The bottom line

DVMT was a 1990s instrument repurposed for a 2010s megadeal. It worked because the cash-equity gap on a $67B tech LBO genuinely couldn't be bridged any other way, and because the asset being tracked (the VMware stake) was high-quality enough that the market would accept a paper claim against it. It cost real money in the form of a persistent trading discount and real exposure in the form of shareholder litigation — but those costs were smaller than the alternative of either walking away from the deal or raising $20B more of cash equity that didn't exist.

The transferable lesson isn't "use tracking stock." It's that the structure of an LBO is often more important than the price. Sometimes the only way to make a deal financeable is to invent a security that converts an asset you'll own into a form of consideration the seller will accept. The principles for doing that — keeping the cash equation balanced, sizing leverage against real EBITDA, exiting when the market lets you — are the same whether the deal is $500M or $67B.

---

Sources:

1. Dell Technologies 13E-3 Filing (2016), SEC EDGAR — DVMT mechanics, deal structure, capital structure detail
2. Wall Street Journal, "Dell to Buy EMC in Record Tech Deal" (Oct 2015), WSJ.com — original transaction announcement, pricing, and structure
3. Silver Lake Q4 2018 Investor Letter (excerpted in Pensions & Investments), P&I — Silver Lake tranche returns and reverse-merger exit detail