← All posts
October 29, 2026·12 min read·LBO · Retail · Capital Structure · Returns

The Michaels Stores LBO: How Bain and Blackstone Took an $800M Pre-IPO Dividend Out of a Peak-Cycle Specialty Retail Deal That Was Supposed to Be the Next Toys R Us

In October 2006 Bain Capital and Blackstone took Michaels Stores private for $6B at 11x LTM EBITDA and 7.8x leverage — the specialty-retail twin of a deal cohort that mostly ended in Chapter 11. Eight years, one $800M pre-IPO dividend recap and five secondary offerings later, the sponsors walked out at roughly 2x the money. Then Apollo took the same company private again.

The 2005-2007 specialty-retail LBO cohort reads like a casualty list. KKR and Bain took Toys R Us private for $6.6B in July 2005; it filed Chapter 11 in September 2017 and liquidated. TPG and Leonard Green took J. Crew private in March 2011 (adjacent vintage, same category); it filed Chapter 11 in May 2020. Apollo took Linens 'n Things private in 2006; liquidation in 2008. Ares and CPPIB bought Neiman Marcus from TPG and Warburg in 2013; Chapter 11 in May 2020. The pattern is so complete that "sponsor-backed specialty retailer bought at a peak multiple" is a diligence red flag.

Michaels Stores belongs on that list — a $6 billion take-private at 11x LTM adjusted EBITDA and roughly 7.8x leverage, signed in June 2006 and closed on October 31, 2006, eighteen months before the housing crisis started chewing through comparable-store sales at every big-box retailer that touched home discretionary. By the structural tells that killed Toys R Us and J. Crew — entry multiple, leverage ratio, category cyclicality, sponsor dividend-recap pressure — Michaels should have been Chapter 11 by 2017.

It wasn't. Bain Capital and Blackstone sold their last share in December 2016, roughly 10.2 years after the close, at a cash multiple that outside analysts estimated at approximately 2x invested capital. Four years later Apollo took the same company private again for $5 billion. Two LBO cycles, same asset, same industry — and no bankruptcy in between.

This is the story of how a specialty retail LBO threaded a cohort that mostly died, and why "survived" is not the same thing as "won."

The deal

Michaels Stores, Inc. was the largest arts-and-crafts specialty retailer in North America — roughly 925 stores and $3.7 billion in pro-forma LTM net sales as of July 29, 2006. On June 30, 2006, after a competitive auction process that reportedly included Kohlberg Kravis Roberts and Clayton, Dubilier & Rice as rival bidders, Michaels agreed to be acquired by affiliates of Bain Capital Partners, LLC and The Blackstone Group L.P. at $44.00 per share in cash. The deal closed on October 31, 2006.

Deal termValue
Signing dateJune 30, 2006
Closing dateOctober 31, 2006
Price per share$44.00 cash
Premium to pre-rumor share price~29%
Transaction value~$6.0B
LTM Jul 2006 pro forma revenue~$3.7B
LTM Jul 2006 pro forma adjusted EBITDA~$541M
Entry multiple (EV / LTM adj. EBITDA)~11.0x
New debt at close~$4.2B
Sponsor equity (Bain + Blackstone + Highfields)~$1.7B
Leverage at close (debt / LTM adj. EBITDA)~7.8x

The capital structure at close was a $3.4 billion senior secured credit facility layered under $1.4 billion of senior, senior subordinated and subordinated discount notes. Four Bain directors and two Blackstone directors joined the board. Highfields Capital Management came in alongside as a smaller co-investor and preserved a board seat.

Two things in that table set up what came next. First is the entry multiple. 11.0x on a specialty-retail business was aggressive for the category — not Hilton-at-21.7x aggressive, but a full turn or more above where good specialty-retail LBOs had historically cleared. Second is the leverage. 7.8x against LTM EBITDA is a thin margin against a business whose cash conversion runs 10-15 points below EBITDA margin once you fund inventory for a seasonal holiday build. The deal was priced for growth, and specifically for the 2004-2006 run of mid-single-digit comps to keep running.

For readers who want to see how an LBO at this leverage ratio balances across the capital stack, the sources & uses walkthrough covers the standard template. To test whether a given debt-service burden survives a stress case, the free debt capacity calculator surfaces which covenant binds first.

What hit it

The arts-and-crafts category is more defensive than toys or apparel — the median customer is a 45-54-year-old suburban woman doing a Halloween wreath or a children's school project, not a shopper trading up on fashion. That mattered when the housing crisis arrived.

In fiscal 2008 (year ended January 31, 2009) Michaels' comparable-store sales fell 4.6% and total net sales fell 1.2%, or $45 million, from fiscal 2007. The weakest quarter was Q3 fiscal 2008, with comps at -6.5%. The decline was concentrated in the floral and home-décor categories most exposed to the housing cycle; the core craft categories held up materially better. By Q3 fiscal 2009 comps were back to +1.3%, and for the first nine months of fiscal 2009 comps were only down 0.5% year over year.

Compare that to the Toys R Us trajectory over the same window: comparable-store sales declined every single year post-LBO, and operating cash flow could not consistently cover roughly $400M of annual interest expense. The sponsor-era balance sheets look superficially similar — both deals were around $6B of transaction value, both carried mid-7x leverage, both closed inside the peak-cycle window. But Toys R Us was being disrupted by Amazon, Walmart and Target on its best-selling SKUs. Michaels was selling scrapbook paper, yarn and silk flowers in a category Amazon never cracked and the general-merchandise discounters have never fully served.

Fiscal 2008 EBITDA still generated enough free cash to service the LBO debt stack. Covenant compliance was tight through 2009 but was never breached. The lenders never got leverage over the sponsors. For a deeper look at the mechanics that put retail LBOs under water when the thesis doesn't hold, see the Toys R Us dividend-recap teardown. Michaels avoided it in 2008-2010 — but then did its own, in 2013.

The pre-IPO dividend recap

In July 2013, with the business stable, same-store sales recovering and credit markets wide open, the sponsors ran the recap. The vehicle was not Michaels Stores, Inc. itself; it was a shell structure above the operating company: Michaels FinCo Holdings, LLC and Michaels FinCo, Inc. On July 29, 2013, FinCo issued $800 million aggregate principal amount of 7.50%/8.25% senior PIK toggle notes due 2018 in a private placement.

Net proceeds — approximately $782 million after discounts and expenses — were distributed from FinCo Holdings to the parent, and the parent paid a cash dividend to its equity holders. The parent's equity holders were Bain Capital, Blackstone and Highfields.

A few mechanical features of a FinCo-level PIK dividend recap are worth noting:

  • Off-balance-sheet to the opco. Michaels Stores, Inc. — the SEC-reporting entity — did not record the PIK notes on its own balance sheet. The debt lived at a parent holding company whose only assets were the equity of Michaels and its subsidiaries.
  • Cash interest serviced via upstream dividend. The opco paid roughly $60 million a year in cash dividends up to Holdings to fund PIK note interest, when FinCo elected to pay in cash rather than toggle to in-kind.
  • Independent monetization event. This recap paid the sponsors $700-plus million of cash before the IPO. Whatever the IPO priced at, the sponsors had already locked in a materially better MOIC than the one from the IPO alone would suggest. The mechanic is the same one covered in the dividend-recap primer: a sponsor return is not the same thing as an exit, and the two don't need to happen on the same day.

The IPO was not the exit

On June 27, 2014, The Michaels Companies, Inc. priced its IPO at $17.00 per share — the bottom of the filed $17-$19 range — and sold 27.8 million primary shares for approximately $472 million in gross proceeds. The shares began trading on Nasdaq under the ticker MIK. None of the gross proceeds went to the sponsors; the company used the IPO proceeds to redeem $466 million of the Holdco PIK notes issued the prior July. Post-IPO, Bain and Blackstone each held approximately 40.1% of the outstanding shares.

The IPO looks like the exit on a transaction timeline, but — exactly as with Realogy, Hilton and most multi-year sponsor sell-downs — the actual sponsor monetization came in a sequence of secondary offerings over the following 2.5 years.

DateTransactionSponsor ownership after
Jun 27, 2014IPO: 27.8M primary shares @ $17.00; $472M to company for Holdco PIK redemptionBain + Blackstone each ~40.1%
Jan 2015Secondary offering of ~18.8M shares at ~$27Combined ownership ~69%
Jul 2015Secondary of 12.5M shares (Goldman block trade)~63%
Mar 29, 2016Secondary of 17M shares + 2M concurrent company repurchase; ~$443M gross~51%
Jul 13, 2016Secondary of 11M shares + 1M concurrent company repurchase~34%
Dec 12, 2016Secondary of 12M shares — final sell-down0%

Between the IPO and December 2016, each of the five follow-on offerings happened at a price above the $17 IPO price. The sponsors sold into a stock that traded at $25-$35 for most of 2015-2016. On a weighted-average basis across the five secondaries, the realized sell-down price was materially higher than the IPO level — a quiet reminder that an IPO-day stock price is a distribution event, not a valuation event, and that a sponsor patient enough to sell down over 2-3 years usually gets paid better than one that dumps at the lockup expiry.

The return

Public filings do not break out a fund-level MOIC for Bain or Blackstone on Michaels. Buyouts Insider's pre-IPO estimate, built at the $18 midpoint of the filed IPO range and including paper gains plus the 2013 PIK-funded dividend, put the sponsors at approximately 2.2x their money. An August 2014 analysis by the Center for Economic and Policy Research, using the $5.8 billion-plus purchase price and the $800 million dividend recap, estimated the realized cash multiple at approximately 2.0x.

The gap between those two numbers — 2.0x versus 2.2x — is primarily the difference between pricing the IPO at $17 (what actually happened) versus $18 (the midpoint), and between CEPR's recap haircut and Buyouts Insider's full-face treatment. Neither number is a sponsor-reported figure. The honest range is "approximately 2x the money."

That return matters for a specific reason that returns-metric posts in this library emphasize over and over: a 2x cash multiple spread over 10.2 years is a sub-hurdle IRR. The arithmetic is roughly 7-8% ungrossed — above the risk-free rate, well below a Fund V or Fund VI's 20% target IRR, and below where a typical LP's cost of capital sits once you adjust for carry and management fees. The LBO returns calculator makes the same point concretely: lengthen the hold on a 2x deal from six years to ten and the IRR drops by roughly 500-600 basis points, with no change in the dollar outcome.

Across the 2005-2007 specialty-retail LBO cohort, 2x over ten years is the top of the distribution. The median outcome in that cohort was closer to a zero or near-zero realized return. Michaels was the one that survived. For LPs it was a positive outcome; for the sponsors it was a mediocre fund-level contribution; for the business it was a decade of operating under a holdco-PIK capital structure that limited reinvestment capacity.

Apollo takes it private again

On March 3, 2021, Apollo Global Management announced an agreement to take The Michaels Companies private for $22.00 per share in cash via a tender offer. The headline numbers: equity value approximately $3.3 billion, total transaction value approximately $5.0 billion including the assumption of ~$1.7 billion of existing debt, premium of 47% to the February 26, 2021 closing price. On April 14, 2021, the tender offer expired with approximately 85.92% of outstanding shares tendered. The second-step merger closed the next day, and MIK was delisted from Nasdaq.

Two cycles, same asset — a template that is rare in large-cap PE. The 2006 and 2021 deals look superficially similar (specialty retailer, mid-single-billions equity check, operational-improvement thesis) but ran on structurally different premises. The 2006 deal was priced for comps growth to continue. The 2021 deal was priced for pandemic-era crafting demand to normalize, for the move to omnichannel to be a defendable moat, and for the retail LBO template under COVID-era interest rates to be viable. By the standard measure of exit quality the second cycle is still unresolved as of this writing.

What the deal teaches

Three lessons.

Entry multiple at peak is not fate. Michaels, Hilton and Dollar General all closed at peak-cycle multiples (11x, 21.7x and ~10x respectively) inside the same 2006-2007 window. All three delivered positive sponsor returns. The three that got wiped — Toys R Us, Harrah's and TXU — did so because of a specific operational or secular factor on top of the entry multiple, not because of the multiple itself. The right way to read a peak-multiple entry is as a question about the quality of the underlying business, not as a conclusion about the deal.

A dividend recap is a return event in its own right, not just a bridge to the exit. The $800 million PIK recap in July 2013 locked in the majority of the sponsors' realized return — before any public-market pricing had been determined. Modeling an LBO as if "exit = fund return" misses the way sophisticated sponsors monetize across the hold. The right template separates (i) running management fees, (ii) recap dividends, (iii) IPO-day distributions, and (iv) secondary-offering distributions as four separate contribution lines in the fund-level return build.

Hold length is the quiet tax on peak-cycle wins. The 10.2-year hold cut the ~2.0x cash multiple down to a 7-8% IRR. Below carry. Below LP hurdle. The cash was returned, but the capital was dead for a decade. Peak-cycle deals that "survive" tend to generate this shape of return — the sponsor doesn't lose money, but the opportunity cost is enormous, and that doesn't show up in the headline multiple that gets quoted at fund-raising dinners.

Read the capital structure before you read the pitch deck. And look at the hold-length column before you let the MOIC column impress you.

Model LBO transactions the way the sponsors do.

The All-in-One PE Model ships with the full LBO stack — sources & uses, debt schedules with covenants, holdco PIK and dividend-recap mechanics, IRR/MOIC returns and sensitivity tables — the same mechanics behind every deal in this library. See the sheet preview.

Sources:

Use the same model on every deal.

26 fully-integrated worksheets. LBO + DCF + Comps + 3-statement + debt schedules + EVA. One-time download.

100% money-back if you find any error
Keep reading
LBO
The Realogy LBO: How Apollo Bought the Country's Largest Residential Brokerage Eight Days After New Century Filed Chapter 11 — and Still Got 2x Back
In April 2007 Apollo took Realogy — Coldwell Banker, Century 21, Sotheby's International Realty, ERA — private for roughly $9 billion. Eight days earlier New Century Financial had filed Chapter 11. Six years, $3 billion in cumulative losses and a sponsor-led debt exchange later, Apollo exited at roughly 2x MOIC.
LBO
The Avaya LBO: How TPG and Silver Lake's $8.3B Enterprise Telecom Bet Filed Chapter 11 Twice in Six Years
In October 2007 TPG and Silver Lake took Avaya private for $8.3B at $17.50 per share. Nine years later the company filed Chapter 11 and wiped the sponsor equity. Five years after that, the reorganized company filed Chapter 11 again and wiped the public shareholders. Inside the slow-motion LBO bust.