# Single-stock daily trading: active paper method v2 Revised 22 October 2026 following the user's instruction to remain invested in the most attractive researched stocks unless a correction or crash is expected. Initial capital: 500,000 DKK. Activated by the user on 22 October 2026 for an initial regular-session paper allocation. Live state, immutable transactions and performance are in `singleStockPortfolio` in the dashboard JSON. These are model observations, not real broker orders. See `single_stock_allocation_2026-10-22.md`. This version supersedes [v1](single_stock_strategy_v1.md). The [initial research](single_stock_research_2026-10-22.md) remains a dated v1 audit. Its all-cash selection result is not the current allocation policy. V2 changes were explicitly requested before model inception; they are not a retroactive change to any trade, price or return. ## Main rule: hold the most attractive eligible stocks The model normally owns a diversified selection of the strongest researched companies. A missing fresh catalyst, absence of a perfect chart pattern or a sideways index does not by itself justify cash. Rank eligible stocks, fill the stock allocation and retain good holdings while they remain competitive. Target eight to ten holdings in ordinary conditions. An existing holding does not have to requalify as a brand-new breakout every day. Cash above the ordinary reserve needs a documented correction/crash case, a binding portfolio risk limit, or a specific temporary implementation problem. The latter must be labelled as a data/execution shortfall, not a market forecast. This is a daily-reviewed stock-selection model with a minimum intended holding period of one trading day. Review at 18:00 Europe/Copenhagen, at most one trade batch per US session. There is no requirement to trade every day or sell after a fixed number of sessions. Holding an attractive company longer is preferable to creating turnover simply to meet the day-trading label. ## Exposure targets | Assessment | Stock target | Permitted stock band | Cash at target | Planned risk per new position | Aggregate planned downside ceiling | | --- | ---: | ---: | ---: | ---: | ---: | | Ordinary conditions | 90% | 80–95% | 10% | 0.75% NAV | 6.0% NAV | | Cautious; no explicit correction base case | 75% | 65–85% | 25% | 0.60% NAV | 4.5% NAV | | Correction expected | 40% | 25–50% | 60% | 0.35% NAV | 2.5% NAV | | Crash/disorderly market expected or occurring | 10% | 0–20% | 90% | 0.20% NAV | 1.0% NAV | These are proposed model settings, not empirically optimal allocation weights. Invest toward the target subject to the mandatory issuer, price, concentration and planned-risk limits. Do not widen a stop or ignore debt/execution problems to hit a target. If a constraint binds, hold the other eligible names, document the precise shortfall, and research alternatives rather than waiting solely for perfect breakout signals. Small deviations from share rounding require no trade. A cautious assessment still keeps most capital in stocks. Ordinary volatility, expensive valuation, price below a 50-day average, VIX at 30, or a missing earnings beat alone does not imply a correction or crash. ## What permits a larger cash allocation Every daily assessment records the state, evidence cutoff, source links, horizon, downside scenario and what would reverse the decision. Forecasts are judgments and can be wrong; never describe a correction as certain. For model classification, a correction case means a central expectation of roughly a 5–15% broad-market decline over the next 0–3 months. The word "lower" alone is insufficient. Require a quantified scenario and at least two distinct corroborating channels from the list below, with current primary evidence supporting the underlying observations: - Broad deterioration: a sustained break of the 200-session average, worsening participation or repeated lower market highs/lows. - Earnings: broad estimate cuts, worsening issuer guidance or margins, rather than one company's disappointment. - Credit/funding: material widening of credit spreads, funding dislocation or a verified banking/liquidity problem. - Macro/policy/supply: a documented shock whose transmission plausibly causes a material earnings or discount-rate reset. A crash case means a credible scenario of at least a 20% broad-market fall, or observed disorderly trading/funding stress. Require documented severe evidence; a volatile trading day or an adverse headline is not enough. Genuine market dysfunction can warrant immediate reduction at the next executable review without waiting for confirmation sessions. ACWI's 20/50/200-session averages and official VIX are evidence, not standalone switches into cash. Retain the existing oil/yield/credit/earnings triggers as context, with their actual persistence conditions. Count distinct risk channels, not multiple retellings of the same story. Enter the correction/crash state only with the documented assessment. For a move back toward higher exposure, require two completed sessions of improving market evidence and an updated explanation of why the downside case has eased. Re-enter ranked stocks in the permitted batch; record missed rebounds as real opportunity costs. If risk data is unavailable, retain the last supported assessment and current holdings unless verified evidence requires an exit; do not infer a crash or fabricate data. ## Current provisional application The existing 21 October outlook says "lower and volatile" with high valuation risk, while earnings and last-verified credit conditions provide counterweights. The 21 October ACWI screen is mixed and VIX closed at 16.68. No quantified 5–15% correction central case or crash assessment was established in the initial selection research. Therefore the provisional v2 planning state is **cautious: 75% stocks / 25% cash**, or 375,000 DKK / 125,000 DKK on the proposed capital. The initial allocation uses this cautious assessment and supported open-session quotes. Keep reassessing the market and ranking the companies. Do not treat the v1 subtotals as v2 scores. If a new, properly supported correction case is established, adopt the defensive target rather than ignoring it to stay invested. ## Mandatory investability and business checks - Established NYSE/Nasdaq ordinary shares and liquid sponsored US-listed ADRs; one listing per issuer; at least five years of listing history. Minimum market capitalisation USD 20 billion, price USD 10, and 20-session mean daily traded value USD 100 million. - The starting research universe is the 38 named companies in the initial audit. Expand it when needed to fill sector gaps, recording additions before scoring. It is not an exhaustive or survivorship-free backtest universe. - Verify mainstream retail-broker availability and the exact listing/currency/ ADR ratio; no CFDs or alternative instruments as silent substitutes. A real entry requires a supported regular-session quote, spread <=0.15%, FX and cash, all with timestamps. - Positive trailing earnings and operating cash flow, positive free cash flow in at least two of the last three fiscal years, and a credible balance sheet. Reject unresolved accounting, solvency or going-concern concerns. - Initially use nonfinancial operating businesses; banks, insurers and REITs need separate scorecards. Payment-processing operating businesses can be assessed on their own financial statements. - No shorting, derivatives, leverage or borrowed money. Missing required execution/business data means skip that issuer and use another verified candidate, not an automatic all-cash allocation. - Annual reports, filings and issuer releases support business facts; identify the provider, horizon and date for estimates. Use one prior completed US session for technical comparisons. Execution prices cannot generate their own signals. Do not mix dividend-adjusted returns with raw-price triggers without translating them onto a consistent share-price scale. ## Ranking: 100 points Rank eligible stocks relative to one another. **There is no 75-point entry gate and no compulsory fresh catalyst.** The purpose is to identify the strongest available businesses at a defensible valuation and risk, not to require every desirable condition simultaneously. | Component | Points | Five-point evidence checks | | --- | ---: | --- | | Quality and moat | 30 | Concrete competitive-advantage mechanism; stable/improving margins on a consistent two-year basis; positive FCF in two of three years; positive trailing net income; ROIC above a sourced estimated cost of capital; net cash or net debt/EBITDA <2.5 with interest coverage >5. | | Valuation | 25 | Trailing P/E <=1.25x the positive comparable-peer median; <=1.25x own comparable five-year median; FCF yield >=peer median; EV/EBITDA <=1.25x comparable-peer median; well-defined forward earnings support growth with nondeclining dated revisions. | | Relative strength and timing | 25 | Close above its 20-session average; above its 50-session average; positive 20-session excess return versus ACWI; positive excess return versus fixed sector proxy; a confirmed breakout or recovery price pattern. | | Catalyst and outlook | 10 | Verified positive operating catalyst with quantified recurring earnings implications; latest issuer outlook intact without a subsequent material adverse development. | | Liquidity and access | 10 | Listing/size/price/history/traded-value screens verified; regular-session spread and exact instrument access verified. | Award five points only for supported tests. Missing information receives zero and a label explaining that it is unverified, rather than false. Show the coverage count separately. A low-information stock is not assumed cheap or high-quality. Lack of a precise NTM consensus series removes that valuation bonus; it does not veto an otherwise eligible, well-understood company. Reported forward P/E with an unknown horizon may be shown for context, not scored as NTM. Use consistent peers and earnings definitions. Moat is a sourced analyst assessment of switching costs, network effects, brands/patents/licences, cost advantages or efficient scale. It is not a proprietary rating unless actually obtained. Low P/E may reflect peak cyclical earnings, one-off gains or deteriorating demand. Cross-check cash-flow yield, debt, stock-based compensation and estimates. Prioritise the highest scores while respecting sector/theme caps and portfolio risk. Break ties using stronger 20-session ACWI-relative return, then ticker. Retain an eligible incumbent in the top twelve unless a feasible replacement scores at least ten points higher. Replacements are reviewed after at least three completed holding sessions; verified risk/thesis exits override that retention buffer. Keep the raw input data and score version for every decision. ## Entry timing and holding policy The original catalyst-continuation and quality-recovery patterns remain useful timing evidence: strong-volume breakouts following published facts, or a recovery near the 20-session average within an established uptrend. Their absence does not bar a core ranked holding. A stock below a moving average can still qualify on business quality and relative attractiveness, with sizing reflecting its actual volatility and downside. Write bear/base/bull cases for each stock, with observed support/resistance, business drivers, adverse evidence and the next relevant event. A plausible 2R trade is a positive timing consideration, not a universal prerequisite for owning the company. Do not invent near-term price targets simply to reach 2R. The current quote can invalidate stale research or require smaller sizing. Keep the strongest holdings rather than imposing the v1 ten-session progress exit or twenty-session maximum. Sell for a broken thesis, verified risk trigger, meaningfully superior replacement, or a change in the market allocation case. The minimum intended hold remains one trading day. Daily review levels cannot guarantee a stop-price fill or cap losses during gaps. Do not open a new position within two US trading sessions of scheduled earnings. An existing position enters the event only if its specific scenario and reduced size are documented: cap event exposure at 5% NAV per issuer and 15% in aggregate, otherwise rotate the excess into the next eligible names. This avoids a blanket sale into cash every earnings season. Unresolved event risk can still require a full exit. Never use a pre-release close as confirmation of after-close news. ## Position sizing and compatible risk budgets - New positions >=10,000 DKK; maximum 10% NAV per issuer, ten issuers, 25% NAV per sector and 30% in one correlated theme. At inception the issuer cap is 50,000 DKK. Whole shares; a subsequent decline below the floor does not force an averaging-down purchase. - Initial review invalidation is entry minus two ATR, where ATR is the simple mean of fourteen completed true ranges. True range is max(high–low, |high–previous close|, |low–previous close|). Use consistent split-adjusted OHLC and current share units. - Size = minimum of 10% NAV, the state's per-entry planned-risk budget divided by the fractional two-ATR distance, available cash and remaining allocation/ risk caps. Round shares down; skip sub-10,000 DKK entries. - Ordinary planned risk is 0.75% NAV (3,750 DKK initially); cautious is 0.60% (3,000 DKK); correction is 0.35% (1,750 DKK); crash is 0.20% (1,000 DKK). For example, a cautious 5% distance implies 60,000 DKK by the risk formula, capped at 50,000 DKK by the issuer rule. - Sum current value x max(0, (current price–invalidation)/current price) over holdings and enforce the state's aggregate ceiling. The cautious ceiling is 22,500 DKK at inception; ordinary is 30,000 DKK. These replace v1's 10,000 DKK aggregate ceiling because the old ceiling conflicts with broad investment at typical stop distances. They are larger loss-planning budgets, not guaranteed maximum losses. Gaps, correlations and FX can exceed them. - Prefer lower-volatility eligible alternatives or smaller positions when aggregate risk binds. Never widen invalidations simply to meet the allocation target. Document target versus actual exposure and the specific constraint. - At most one batch per US session. The initial funding/re-entry batch may build toward the assessed target. Subsequent ordinary replacement purchases total at most 25% NAV per session; verified risk exits are uncapped. Normal orders >=10,000 DKK; full exits may be smaller. Avoid cosmetic rebalancing. - At a 6% portfolio drawdown, require a documented review of ranking, correlated exposures, thesis quality and risk limits at the next run. Reduce broken positions and rotate into stronger eligible alternatives; do not declare a market crash or liquidate all stocks solely from portfolio underperformance. ## Daily exits, accounting and evaluation At each executable review, exit a breached prior invalidation or an invalidated business thesis and rotate/reduce according to the allocation and ranking rules. Never widen a review level to avoid acknowledging a loss. Trail upward using the larger of its prior value and the highest completed close since entry minus two current ATR. Use only information available before the decision. These are once-daily review triggers, not resting broker orders. Intraday crossings are not retroactively filled at the trigger. Use the actual supported quote at the next permitted review; preserve real gap losses. On holidays, early closes before the run or unsupported quotes, make no orders and explain the execution shortfall. Existing holdings continue to be marked using supported observations with their true session dates. This remains a separate 500,000 DKK model. On activation record actual prices, FX, shares and cash plus an identical-interval ACWI DKK benchmark. No backdated fills. Show percentage performance with at least +/-10% chart range, each position's return, realised/unrealised P&L, distributions, exposure, turnover, drawdown, ranking changes and a reasoned trade journal. Handle cash/ADR distributions and FX without double counting. Primary P&L excludes commissions and tax as in the existing models; show the separate illustrative 0.10%-per-trade and 0.25%-per-actual-FX-conversion sensitivity from v1. Keep the currency cash ledger and identify midpoint versus bid/ask marks. These assumed costs are not broker quotes. Retain every score, rejected candidate, old trade and allocation rationale. Record a reason for cash above the relevant band. Evaluate the record over at least 60 sessions and 30 closed trades before provisional conclusions; it is still a small sample. The model was activated on 22 October 2026. The initial user-authorised open-session batch counts as that session's only batch; the 18:00 job may revalue/research it but cannot trade it again that day. ## Research basis and limitations The evidence supports evaluating quality, price trends and cash opportunity cost; it does not validate these exact allocation targets, scores or short holding periods. Choosing an exposure mandate is a design decision and increases losses when equities fall. In particular, failure to forecast a correction is not evidence that no correction can occur. - [Vanguard cash-allocation framework](https://www.vanguard.ca/en/insights/framework-for-allocating-to-cash): excess cash/market-timing trade-offs, with explicit attention to risk tolerance and horizon; not proof of short-term stock-trading alpha. - [AQR Quality Minus Junk](https://www.aqr.com/Insights/Research/Working-Paper/Quality-Minus-Junk): business quality evidence at a different horizon and long/short design. - [MSCI Momentum Indexes](https://www.msci.com/indexes/group/momentum-indexes): institutional momentum framework, not validation of daily entries. - [Barber et al. day-trading study](https://faculty.haas.berkeley.edu/odean/papers/Day%20Traders/The%20Cross-Section%20of%20Speculator%20Skill.pdf): historical difficulty of consistent after-fee intraday profits. - [SEC day-trading discussion](https://www.sec.gov/about/reports-publications/investorpubsdaytipshtm): monitoring, cost and loss considerations. - [Liquidity and earnings drift](https://business.columbia.edu/faculty/research/liquidity-and-post-earnings-announcement-drift): counterevidence on transferring earnings-drift profits to liquid stocks. - [Investor.gov order types](https://www.investor.gov/introduction-investing/investing-basics/how-stock-markets-work/types-orders): execution-price limitations.