Actionable insights — Don't chase yield, do this instead

7:25 1. Set a two-part income target: yield and growth

1. Set a portfolio yield floor relative to the market — about twice the S&P 500's yield.
2. Require dividend-per-share growth above inflation, so income rises in real terms.
3. Reject names that meet only one leg: a high static yield without growth, or a growth stock paying a token yield.

• The portfolio's yield versus the S&P's; year-over-year dividend-per-share growth versus CPI.

10:51 2. Balance sheet first, then free-cash-flow coverage

1. Check leverage before margins or returns: a lightly levered company keeps the flexibility to survive a cycle or fight a disruptor.
2. Test coverage on cash, not earnings — "dividends are paid with cash": operating cash flow minus capex, versus dividends paid.
3. Normalize for working-capital swings so one good or bad year doesn't distort coverage.

• Net debt / EBITDA rising; FCF payout ratio drifting toward or above 100%; capex rising faster than operating cash flow.

12:35 3. Flag dividends raised on a boom

1. Look for outsized dividend increases (well above the company's history) made during a cyclical upswing.
2. Ask whether the earnings that funded the raise were a new normal or a temporary spike.
3. Stress-test: can the new dividend be paid in 10 years at a normal earnings growth rate? If not, expect token raises or a cut.

• Year-over-year dividend hikes far above trend; earnings falling back after a demand shock; labour-contract renewals; raises shrinking to a penny.

14:33 4. Read an abnormal yield as the market's cut forecast

1. Compare the current yield with the company's own normal range.
2. When a usual 3–4% payer climbs to 5–7% without faster growth, assume the market is pricing a cut until the balance sheet and coverage prove otherwise.
3. Don't use a high yield as a timing tool in sectors facing structural disruption; it worked when regular bull/bear cycles pulled yields back down, less so now.

• Yield versus its 5- or 10-year range; sector-wide dividend cuts; volume declines in the category.

16:25 5. The fallen-angel screen

1. Screen for companies with long records and dividend growth that are down 10–40%.
2. Check the earnings revisions: if estimates "are barely coming down," the selling is impatience or a theme (not AI-related, feared AI disruption), not fundamentals.
3. Then ask whether the feared disruption really changes the business model before buying.

• Price drawdown versus the change in forward EPS estimates; dividend raises continuing through the sell-off; evidence of AI hurting or helping pricing and retention.

18:48 6. Find what the sector's multiple assumes — then test it

1. Identify the cheapest sector and state the assumption that makes it cheap.
2. List the physical evidence that the assumption is too easy (inventories, reserves, underinvestment, policy shifts).
3. Express it through the companies that sell into the rebuild, not only the commodity producers.

• SPR refill announcements; OECD inventories; international upstream capex budgets; Venezuela production.

26:05 7. Asymmetric sell discipline

1. Handle losers mechanically with a stop-loss, so a broken stock doesn't become a big mistake.
2. Don't sell winners on valuation alone: if the thesis hasn't changed and the business keeps strengthening, it may deserve a permanently higher P/E.

• Whether a rich multiple comes with rising returns on capital and dividend growth (hold) or with slowing fundamentals (trim).