Video summary

$100M Trading Coach: I Fire Traders Who Make Money If They Make THIS Mistake!

Main summary

Key takeaways

Business

Business / execution focus (prop desk strategy & operating model)

Jeff Holden (SMB Capital, head of trader development) frames trading as an execution-and-system business: the trader’s job is to create and execute positive expected value (EV) per decision, using a playbook where structural factors come before “psychology.”

Core operating thesis

  • Expected value first: “Your job… is to put yourself in the opportunity to have a positive expected value for every decision.”

  • Structure before psychology

    • Psychology is often not the root cause; the system/playbook is.
    • Example: a trader who “sells half early” attributes it to psychology, but the issue may be broken rules/position management, even if they “feel” psychologically controlled.

“Player vs Player” (PVP) and when technicals matter

  • Technicals = price action structure (not “indicator magic”).
  • Technicals’ utility depends on the environment:
    • Shorter timeframes → more PVP participation and more immediate reaction → technicals matter more.
    • Fundamental catalyst / narrative shift → technicals matter less initially (price moves because participants change behavior).
    • Overextension / range breaks → technicals matter again because participants must act at key levels.

Frameworks / playbooks explicitly discussed

1) Desk “Playbook” concept: EV + structured story

Build an institutional-grade narrative:

  • Catalyst: why something is changing
  • Setup: higher-timeframe “structure” / technical condition
  • Trade / Tape: how participants are expressing it now; expected speed/magnitude

Edge depends on knowing when the “story” is valid vs when it breaks.

2) “ASSET” protocol (risk-first trade design)

A minimum structure for every trade idea:

  • A = Allocation

    • Decide risk per trade before thinking about entry.
    • Starting model: account size ÷ 50 = daily stop number (≈ 2% daily), then not all of it per trade.
    • After ~20 reps, allow dynamic risk allocation by setup probability + R:R (not just rules-of-thumb).

  • S = Stop

    • Define stop first, then size from allocation.

  • E = Entry

    • Entry must align with the plan (predictive levels vs reactionary confirmation vs scaling rules).

  • T = Time / Target

    • Use standardized targets (e.g., measured moves / opening range break).
    • Example: opening range high/low → target based on 2 measured moves (range high→low, then add).

3) Trade selection logic: “Probability + Risk/Reward = EV”

  • Not just win rate: define probability as the likelihood the trade works, then pair it with R:R to estimate EV.
  • Starting guidance (risk model):
    • Aim for roughly 3:1 reward-to-risk with about 50% probability.
    • Rationale: people often overestimate probability and underestimate true R:R; balancing the equation improves results over time.
  • Operational point: you must understand why the edge exists so you can detect when it disappears (edge is not permanent).

4) Learning system: reps + trade write-ups (iterative development loop)

  • For each strategy:
    • Trade it ~20 times to accumulate enough reps for feedback (some quantitative requirements may be higher, but operationally 20 is “enough” to learn).
  • The process is “write it up to learn,” not “write it up to be perfect.”
  • Post-trade questions that matter:
    • What happened? Why did it happen? What can I learn?
    • Avoid: “I was frustrated,” “market screwed me,” etc.

5) Skill-specific training “touchpoints” (weekly operating rhythm)

Recurring coaching loops include:

  • Weekly meetings:
    • Reading the tape meeting
    • Technical analysis meeting
    • Top ops meeting (share opportunities traders found; helps build playbooks)

Targets for structure:

  • ~50 touchpoints/week per trader
    • includes daily report cards reviewed with feedback

Accountability + learning:

  • Peer review accelerates learning, but is less “discrete” than mentor feedback.

Concrete examples & applied recommendations

Example: why “technicals don’t matter” is situational

  • Semiconductors example:
    • Rally not from technical breakout, but from a fundamental shift in business transactions (e.g., a change in memory demand narrative).
  • Overextension / range break:
    • Technical structure matters when participants are forced into actions at key levels.

Example: channel strategy edge (mean reversion / self-fulfilling)

  • Channeling (support/resistance bands respecting “channel lines”):
    • Edge can come from participants treating it as pseudo-trend/mean reversion.
    • News/volume can break the channel → that’s a different system.

Example: head-and-shoulders speed of response defines edge

  • Edge is not the pattern shape; it’s the expected response after neckline break.
  • Response should be fast because the “battle” concentrates around the neckline area (longs/liquidations/short reactions).
  • Clarifying “wrongness”:
    • Traders aren’t “wrong” at the last high by shape—they’re wrong when the lower high + roll-over invalidates their expectation.

Example: “spaghetti against a wall” failure mode

  • Common beginner bottleneck:
    • No structured trade plan → random attempts.
  • Fix:
    • Require an asset protocol for every idea (Allocation first).

Metrics / KPIs / numerical targets mentioned (execution-focused)

No direct firm financials or trader P&L targets were provided, but operational parameters and model constants were stated:

  • Expected value (EV): central KPI concept (positive EV per decision)
  • Starting trade profile
    • 3:1 reward-to-risk
    • 50/50 probability as a base learning target
  • Daily loss/risk budget model
    • account size ÷ 50 = daily stop number (≈ 2% of account/day)
    • then bet only a subset per trade (because multiple trades/day)
  • Strategy testing
    • ~20 trades as a rep threshold for strategy learning cycles
    • mentions they “quantitatively require 35,” but desk experience says 20 is enough
  • Training cadence
    • ~50 touchpoints/week per trader
    • daily report cards with supervisor/mentor feedback

Leadership / organization tactics (how SMB Capital runs development)

Hiring / scaling rule

During interviews they use two explicit go/no-go questions:

  • Will the person be a positive addition to the firm (make it better)?
  • Will the person be a better person because of the firm (10-year commitment implied as part of evaluation)?

Culture design: individuality with alignment (“pack hunting”)

  • Default: traders pursue unique discrete edges
    • they estimate 80–90% of the time traders trade unique edges
  • Rare moments: “pack hunting”
    • when “everything lines up” and the desk converges on the same opportunity
  • Risk: pack hunting too often wastes opportunities and reduces development of unique edges.

Progress measurement

Progress is defined as number of bottlenecks worked through, not “positive psychology loops.”

  • Example bottlenecks:
    • Missing core discrete risk rules
    • Exiting too early (e.g., selling half immediately) due to rule/process mismatch

Enforcement / discipline

  • Yolo trades that violate standards are treated as disrespect to the firm/capital/team.
  • They’d rather teach and correct than punish, but repeat violations lead to termination.

Investing/markets angle (high-level only)

  • Discusses “prop/funded account” ecosystem at a high level:
    • Funded challenges can become gamified (cheap entry, reset button), potentially allowing profit without true positive expectancy.
    • Professional prop firms are framed as environments to build real edge + accountability rather than chase payouts.

Presenters / sources

  • Jeff Holden — Head of Trader Development, SMB Capital (SMB Capital referenced as the firm behind consistently profitable traders)

  • Video hosts/speakers (unspecified by name in subtitles) include:

    • “Titans of Tomorrow” show host(s) / interviewer (name not stated in the provided subtitles)
  • Co-authored psychologist referenced (name not fully shown): Dr. Seamberger (appears referenced as co-author of a chapter; exact full name not present in subtitles)

Original video