Chief Investment Officer of Oxbow Advisors (Ted Oakley's firm) — value-discipline, downside-first portfolio manager; trims momentum spikes, buys unloved sectors, and keeps clients in short-duration Treasuries while the AI trade runs, with a running synthesis and per-appearance breakdowns.
Recent long-term-growth add — an "industrial-style" niche-products business (vet diagnostics): critical products that are a low share of customer cost, so reliability + pricing power.
Illustrates the opportunistic 10% sleeve — bought its convertible preferred for yield with the common beaten down (buy-low, get paid to wait, convert higher).
Bought Jan >20x FCF when cybersecurity was feared AI-disrupted; doubled to >40x FCF and cut in half — still likes it long-term, valuation now full (trimmed).
Avoid — 20-70x revenue like SpaceX/OpenAI; Oxbow doesn't short or chase momentum, so it waits. Frontier-lab economics also pressured by cheaper Chinese models.
Avoid — too cyclical, thin moat; a bull analyst's own numbers (EPS $250 in 2028 to $50 by 2030) imply ~$400 at a normal 8x vs a ~$1,200 peak = a 2/3 base-case drop.
Avoid — one of three mega private IPOs at 20-70x revenue vs Google's 8.5x; all optimism priced in. Also a source of the one-time gains flattering hyperscaler earnings.
Avoid — 58-79x revenue IPO, a Buffett "too-hard pile"; a bullish sell-side report admits no free cash flow until 2035. Down ~16% from its first-day close.
In one line: A downside-first value manager: reject any new buy with more than ~20% projected downside, trim into momentum spikes and re-enter unloved assets at pre-set targets, hold ~40% short Treasuries while the AI trade runs, and pre-position in the sectors the money must rotate into — bracing for a possible deeper decline in 2027.
Downside-first screen. Before upside, he models a name's price in a normal 20–25% index bear market; anything with >20% projected downside is a pass. The 12–15 highest-quality semiconductors screen to ~40% downside — so the whole AI/semiconductor complex (SOX +230% in 14 months, a dot-com analog) is avoided, incl. Micron, whose bull analyst's own numbers imply a 2/3 drop.
Trim the spike, re-enter at targets. Keep the strategic allocation fixed and treat "hot" as a sell signal: trimmed gold ~$5,000+/silver ~120 and energy on the Iran-war oil spike, now re-adding at gold ~$4,000/silver ~$60 and to E&Ps/refiners/pipelines/oil-services below $70/bbl on a "raised floor." Fortinet was the trim-the-double in action (bought >20× FCF, cut at >40×).
Pre-position the rotation destination. ~60% of trading is momentum money that must stay invested; when it sells one trade it rotates. He buys the forgotten sectors early — industrials/healthcare/financials (IDEXX, McKesson, US Bancorp) and energy/royalties (Northern Oil, Kimbell) — to be "already there" when momentum flips (financials, healthcare, utilities, staples).
60/40-with-short-Treasuries, moving to 30/30/30/10. The stock book runs ~60% equities / ~40% short Treasuries; the target mix is 30% short high-quality fixed income (≤3yr), 30% commodities, 30% high-quality stocks, 10% opportunistic special situations (e.g. a Boeing convertible preferred). Locked more 2-yr Treasuries >4%; long bonds are a structural avoid since 2020.
AI-trade avoidance & IPO patience. Never buy the mega-IPO in year one — cap ~10× revenue for a great grower (Google IPO'd at 8.5×); 20–70× (SpaceX, OpenAI, Anthropic) prices the perfect outcome, and hot IPOs usually halve within a year. SpaceX also fails the free-cash-flow test (no FCF until 2035 per a bullish report he cites). Bought Airbnb only after 4–5 years of patience.
Late-stage warnings & the 2027 watch. ~45% of the S&P is one trade, households ~73% in stocks (highest near retirement), margin-debt/money-supply at records. Base effects should pull inflation toward 2% by spring, but cycling against this year's high-growth/high-inflation first half makes 2027 a candidate for a deeper decline.
Appearances
One dated page per appearance — each has its stock/name table, talking points, and the saved transcript. Newest first.