Winston Zhu

I went through 20 session decks and five days of notes from EWMBA 236M.2. These are the lessons I wish I had at the start.

If I had to boil the course down to one idea, it would be this: a turnaround is mostly about getting people to accept drastic change, then executing the same basic sequence with discipline. Diagnose how bad things are and why. Secure a real mandate. Find the viable core. Build a cash bridge. Write the Recovery Investment Thesis (RIT). Close the gap through volume, margin, and structured cost work. Fix leadership and talent. Track a handful of forward-looking KPIs tied to pay. And always keep a Plan B.

Ten laws across the cases

The turnaround starts before you walk in. The setup work matters as much as the rescue itself (Lyric, Scotts, Hertz, Mercury).
Get the mandate in writing, and know how much leash you have. A good plan is useless if someone can stop you halfway through (Lyric, Tata, HRC, Koss 2.0).
Find the viable core, then starve or sell everything else - even customers. Focus beats sentimentality (Koss, Spansion, Lyric, Burger King).
Cash comes first. Revenue is vanity, margin is sanity, and cash is king and queen. Accounting profit will not fund the next payroll (Scotts, Mercury, HRC, EDS, Cash is King).
Revenue mania kills. Growth that burns cash or destroys margin is not progress (EDS, DRC, Mercury).
The incumbent is usually part of the problem, and incentives are usually why. The course put the incumbent factor at 94.7% (HRC, EDS, DRC, Hertz).
Buy-in is engineered. Make the plan feel achievable, create early wins, use personal scorecards, and do not pretend fear has no role (Tata, Mercury, Burger King, Spansion).
Leverage cuts both ways. Debt can amplify a good plan, but it can also hand the company to someone else (Marvel, Koss, DRC, Kidder, post-PE Hertz).
In bankruptcy, valuation is negotiation. Know the waterfall and buy the fulcrum security (Hawker, Marvel, Spansion, Levin).
Leaders emerge in a crisis, and speed beats perfection. Ready, fire, aim (Chile, Spansion, Burger King).

Case notes

Here is the evidence behind those laws, in course order.

Day 1: Frameworks

1. Lyric Dinner Theatre - Butts in seats buying more booze

Your leverage is highest before you take the job. Deborah fixed operations, repaid the note, and still got fired because she never had a real mandate.
Hit shows drove the economics. Annie produced $1.2M in revenue and $500K in profit, while Arsenic produced $553K and $100K. The lesson was simple: identify the one profit engine and organize around it.
Fixing operations did not answer the bigger question: was this a business or a lifestyle activity? Without that answer, the turnaround was only temporary.
"Vision without execution is hallucination."

2. HRC, the nonprofit rehab center - No margin, no mission

The center lost $605,718, burned through 41% of its cash in a year, had not renegotiated insurance in 17 years, and had a board that "didn't react until they could see the bottom of the cash drawer." The incumbent leader and complacent board were the root cause - a pattern the course said shows up in 94.7% of turnarounds.
Revenue quality mattered more than volume. Cash-pay clients brought in about $500 a day versus roughly $367 from insurance. The plan was to renegotiate payers and reposition the center as a premium option, with prices up 30% and a $350K renovation.
The sequence mattered: preserve cash in days 1 to 30, then fund growth after day 120. The board's job was to govern, not manage.
HRC relapsed in 2017. Turnarounds decay, so early-warning monitoring has to be permanent.

3. Scotts, the CD&R LBO from ITT - Any fool with money can buy a business

The value-creation plan started before the purchase: Targeted, Tooled, Talented. Never buy a company you cannot improve.
Working capital was the deal. Inventory turns improved from 2.08 to 3.20, working capital fell 42% or $25M, and Scotts repaid about $35M of debt early. That cash funded Miracle-Gro and entry into the professional lawn-care market, which later produced about 40% of profit.
When someone has to go, move quickly and treat them decently. In the Joe Turlock role play: say it in the first sentence, do it on Friday, have a witness, and have the successor ready.
Keep the KPI list to five or six forward-looking measures and tie them to compensation. Measurement mania is not management.

4. Hertz, the $15.2B Ford carve-out in 2005 - What must we believe?

The turnaround really began about ten years before the bid. CD&R had the asset-backed-securities banks locked up on retainer, which became a bidding advantage, and its investment committee met ten times.
Benchmarks made the diagnosis concrete: depreciation was 29.9% of revenue at Hertz versus 24.4% at Avis, and capex was 4.1% versus 1.6%. CD&R also hired Avis's former CFO and followed the rule of "two sources for everything."
The root cause was incentives. People were paid for revenue growth rather than profit. The new KPIs - Spread Value, NPS, and ROIC - were tied to bonuses. The result was a 33% IRR and 2.64x return.
The exit came in layers: a $1B dividend recap, an IPO, secondaries, and block sales. The caution came later. Post-PE debt grew from $407M to $18.9B, and Dollar Thrifty became a winner's curse. "Adult supervision is elusive when you diminish your ownership."

Day 2: Cash is King

5. Mercury, the aerospace roll-up - Roll-ups without integration are one hand clapping

A 232-day cash conversion cycle versus 64 days for peers told the story. The fix was a supplier czar, centralization, AR at 30 days, AP at 45, and one ERP. The model showed $233M of cash; actual EBITDA rose 70% in two years, worth roughly $676M.
Integration has to be part of the investment case. Put the top dealmaker on the hook and use a standing integration team.
Buy-in often comes down to personal math. Plant managers were shown what the plan could do for their equity, and they "ended up very rich."
Never walk into a meeting without the votes. Have the meeting before the meeting.

6. Cash is King with Stephan Thomas - Profit is an opinion, cash is reality

The top 1,000 US companies had $1.7T trapped in working capital, equal to 11% of revenue. Cutting DSO from 60 to 50 days on $3.6B of revenue would free $100M.
Working capital is usually the cheapest, fastest source of funding and a good proxy for management quality. You cannot manage what you do not measure, and the 80/20 rule still applies.
Every lever has a trade-off: collections versus customer relationships, inventory versus service. "Working capital is not like good wine: it does not age well."

7. Tata Tea and Tetley - Turning around a failed integration

"Synergy" is a BS word unless it points to specific operating improvements, owners, and deadlines.
Homi first secured a mandate from Ratan Tata, then turned the banks from enemies into allies: restructure the debt, give him six months, and in return the banks would scare both management teams. Fear was a legitimate motivator: "We are like teabags..."
Each stakeholder needed different handling - banks, growers, and Tetley management. The team used a bonsai-tree integration and priced tea at auction value to force a quality focus.
The Kidder/GE parallel was blunt: GE's $600M acquisition died in integration, not valuation.

8. Koss in Chapter 11, 1984 - Viable core meets founder syndrome

Koss made great headphones but picked the wrong market. "We can compete with Sony" was the flawed premise. Koss 2.0 focused on B2B niches such as studios, broadcasters, and schools, cut COGS from 77% to 55%, reduced DSO from 70 to 40 days, and underwrote roughly a 62% IRR.
"Was it stupid to loan a $15M company $13M? Yes. But is it stupid to want it back? No." A lender's willingness is not a reason to borrow.
A bridge needs several spans: equity, borrowing, internally generated cash, and time.
Weak controls hurt twice: first in the 1984 collapse, then in a $34M embezzlement against average net income of about $6M. The financial statements showed the classic pre-failure patterns.

Day 3: Recovery Leadership and Workouts

9. DRC, the repair depot - Is hope a strategy?

The thesis broke when phones became cheaper to replace than repair. You cannot operate your way out of dying economics. DRC switched from repair to salvage and sold while buyers still believed, getting $70M gross from Solectron weeks before the market bottom. Only $23.72 was left in the account.
The pathology was willful optimism: fire the CFOs who brought bad news, then miss EBITDA every quarter. The cumulative miss was $28.5M, a 332% forecasting error.
The IPO failed all four prerequisites, and Goldman canvassed about 30 PE firms without finding a single interested buyer. Know which options are real.
Growth plus leverage plus working-capital intensity equals insolvency. The capstone called this combination out by name.

10. A&M Change Readiness with Will Lovis - Score the organization, not the plan

CEO involvement was the number-one readiness factor. An absent CEO outweighed everything else; one contact-center company never mobilized and missed the AI window.
Diligence behavior predicted post-close behavior. Test the organization with small projects, read exit interviews, and ask people what they disagree with.
Resistance followed organizational impact, not logic. Do the sidebars before every meeting.
The decision was not automatically yes or no: invest, invest with conditions, or pass.

11. Chilean Mine Rescue - Leaders emerge in crisis

Pinera took charge, even if that meant breaking the law; framed the mission honestly but hopefully - "home, dead or alive, sparing no expense"; triangulated to the right leader in Sougarret; brought in experts at any cost; then got out of the way and controlled his impulses.
Crisis work is teaming: teamwork on the fly across boundaries. Relying only on stable teams would have been a mistake.
Execution became a learning loop: ready, fire, aim. Plans A, B, and C ran in parallel, and the problem was split into finding, sustaining, and extracting the miners.
All 33 miners came out from 2,000 feet underground in 70 days, against odds below 1%.

12. Marvel, Perelman versus Icahn - Leverage cuts both ways

Never borrow against your own stock. Pledging 80% of it effectively handed the company to creditors.
Valuation drove the negotiation. Competing 2001 EBITDA estimates of $78.7M and $344.8M produced plan values of $1.1B and $2.2B, with recoveries of 7% and 17%.
While two egos fought, the operator with a real plan - Perlmutter - took the company. Disney paid $4B in 2009 because the viable core, Marvel's IP, had been real all along.
Be careful with owners who already made their exit. At that point, they are playing for "icing on the cake."

Day 4: Bankruptcy and the Turnaround Plan

13. Distressed Debt with Mark Levin - The machinery

Early warnings included bonds at 80 or below, yields to maturity of 10% to 15%, covenant breaches, vendors pulling credit, and widening CDS spreads.
The absolute-priority waterfall shaped everyone's behavior. Buy the fulcrum security where your valuation says the value breaks.
Chapter 11 mechanics included DIP financing, a 12-week budget, 180 days of exclusivity, approval from two-thirds of value and more than half of holders, and cramdown. A pre-pack was the fastest route.
Dislocations create forced sellers. During COVID, CCC yields rose to about 20% versus a normal 7%. Hertz, with $19B, filed while Avis, with $20B, did not. The difference was capital structure, not operations.

14. Hawker courtroom simulation - Valuation is where you sit

Each stakeholder's number matched its place in the waterfall: secured lenders argued for $1.7B, unsecured creditors $2.1B, subordinated creditors plus the PBGC $2.9B, and equity $3.3B. The judge tested $1.7B and $2.3B.
"The more secured you are, the less the business is worth." Bank valuations moved with the client paying for them.
DIP lending offered fees plus first priority: a 2% commitment fee and 1% exit fee. Always ask for fees.
Critical sole-source vendors got cash immediately. Pensions and the PBGC were negotiating parties, not line items.

15. Spansion with John Kispert - The U-turn

The pivot required firing customers. Spansion cut its customer list from 15,000 to 10,000 and dropped Nokia, its largest customer: "We are done, get over it." If Nokia wanted to sue, it could get in line. Embedded memory replaced commodity cellphone memory as the focus.
EBITDA moved from negative $400M in 2009 to positive $400M in 2015. Market cap reached $2.5B, and employees owned 15% of the equity.
"In bankruptcy you are not finding value, you are negotiating value for your vision." The court fight ranged from $671M to $3.2B.
Retention was engineered empathy: pay for the person's real need, sell the challenge, delay the answer, and have dinner with the spouse. Communicate everything 600 times. Do what you say you will do.
"Be decisive: the middle of the road is full of yellow lines and dead snakes."

16. Burger King with TPG and Asiff Hirji - Plan, people, measurement

When banks would not lend, seller financing made the deal possible: $750M of covenant-free Diageo debt plus $213M of PIK. "Sellers are liars unless they are owners."
The 100-day plan was specific: $100M of cash in FY05 and 10% sales growth. Fourteen of the top 15 executives were replaced, with TPG interim leaders filling the gaps.
The team proved the model instead of arguing about it. Company-store pilots persuaded resistant franchisees, saved 1,300 stores, closed 200, and eventually made the NFA disband itself.
The core customer came first. Cutting six seconds from drive-through time added about 1% to sales, or roughly $60M. The Whopper carried an 85% gross margin, while the healthy-menu push flopped. Moving into adjacent markets is not a turnaround.
"Patience is overrated." Turnaround managers often make poor peacetime managers, so know when to leave.

Day 5: Leadership in Practice

17. EDS board meeting - The board was a cause, not a bystander

Revenue mania produced the $6.9B NMCI contract at a 4% margin over the CFO's objection. Free cash flow fell to negative $1.87B, the stock dropped 53% in one day, and the CEO was paid $55M that year. It was a shoot-the-messenger culture.
The warnings had been visible for years. Gross margin fell from 32.2% to 25.7% between 1993 and 1998, while accounts receivable grew at a 20% CAGR versus 15% for sales.
The fix was approval gates for total contract value above $1B, upfront capex above $100M, collections deferred more than 12 months, or margins below 10%. The board had to own the metrics, screen CEO candidates for operating ability rather than dealmaking, and pay salespeople on contract profitability. That is what CD&R's Michael Jordan later did before EDS was sold to HP.
"Govern or get out."

18. Capstone, Session 20 - The toolkit

The RIT had five buckets: the general case; attractions and risks; sized improvements; impact on returns; and what we had to believe. Those fed a five-year exit and IRR. If too many risks were uncontrollable, the answer was simple: run, do not walk.
The best turnaround is the one that never has to happen. Watch disruption, stay close to the core competence, prefer profitable growth, and remove weak leadership quickly.
The slogans held up across the course: "Revenue is vanity. Margin is sanity. Cash is king and queen." "Your ego is not your amigo." "1. TALENT 2. TALENT 3. TALENT."
Private-equity entry multiples rose from about 9x in 2008 to 11x or 12x and above. CD&R still bought around 10x versus a 13x market by partnering with seller-owners. At Hussmann, those owners doubled their money compared with an outright sale. Discipline beat dry powder.