Day Trading
High Probability Trading
Link argues that you already know enough setups and are still losing because you size wrong, trade too often, and never wrote anything down. He answers with actual numbers: half your capital in reserve, 5 percent risk ceiling, a position table built from average true range, and a 45-minute clock on dead trades.
The one idea
Most trading books sell you an entry. Link sells you a filter. His claim is that the difference between the 10 percent who make money and the 90 percent who do not is almost never the signal. It is that the winners have written rules for how much to risk, what a trade has to look like before it qualifies, and when to walk away. The losers have opinions.
The organizing test is a question he wants you to ask out loud before every entry: why am I making this trade? He then prints ten answers that qualify (the stock is stronger than the market and the sector is working, the trend is up and price just pulled back to the moving average, there was bad news and it will not go down, my system fired, it broke a major level with room left) and thirteen that disqualify (I want to make money, I am bored, the market is open, I have extra margin available, I do not want to miss the move, my broker recommended it). If your reason is on the second list, there is no trade. That sounds obvious until you honestly audit a week of your own fills.
Everything else in the book is machinery for making that filter enforceable when you are actually watching money move. The risk parameters get written before the market opens. The maximum position size per market gets computed in a spreadsheet, not decided on the fly. The nightly game plan names the specific levels you want. By the time prices are live, the only thing left to do is execute or pass, and passing is a real option: zero is an acceptable position size.
What it actually teaches
Link's process runs backward from the usual order, setting the defensive numbers first and only then letting a trade qualify.
SPLIT THE ACCOUNT IN HALF
Total capital is not trading capital. Take half, call it at-risk capital, and park the rest in an interest-bearing account you never touch. His logic is arithmetic: if you never have more than 50 percent exposed, a total wipeout is structurally impossible and a full losing streak still leaves you a second career. Every percentage below is a percentage of the at-risk half, not the total. He sets the floor for total capital at $25,000 to $50,000 for futures and $100,000 for equities, and says $5,000 is simply too small to survive a normal string of losers.
CAP RISK AT 5 PERCENT PER TRADE, AIM FOR 2
Fixed-fraction sizing. Five percent of at-risk capital is the outer limit, chosen because it lets you be wrong 20 times in a row before you are finished. Professionals run under 2 percent. On $25,000 at-risk, 5 percent is $1,250. Link is honest that at 2 percent most retail accounts cannot put on meaningful size, and that anyone trading a $3,000 account is risking over 20 percent per trade whether they admit it or not. He also caps total open risk: no more than half of at-risk capital exposed at once, under 30 percent preferred, and correlated positions counted as one (5 percent in crude plus 5 percent in heating oil is a 10 percent bet, so allow maybe 7.5 percent across a whole sector).
BUILD A MAX POSITION TABLE FROM ATR
Divide your dollar risk per trade by the dollar value of the market's 14-day average true range. That number is your ceiling in that market, not your normal size. In the book's Table 15-1, $1,250 of risk against a $4,800 ATR on the full S&P 500 returns zero contracts, the mini S&P at $1,000 returns one, corn at $175 returns seven, and MSFT at a $2.00 daily range returns 600 shares. Ranges change, so the table gets rebuilt. Note these are 2002 volatility numbers.
WRITE THE PLAN, THEN WRITE TONIGHT'S GAME PLAN
Two documents. The trading plan (system rules, money management parameters, which markets, hold times, costs, how you review) changes rarely and should read well enough that a stranger could trade your account from it. The game plan is written every night: what you want to buy, what you want to short, the exact levels, and where you would be out. Link rewrites it at lunch if the day has changed. The principle is that any decision that can be made away from live prices gets made away from live prices.
MAKE THE TRADE ANSWER THE QUESTION
Before entry, ask why am I making this trade. Good answers describe a condition in the market. Bad answers describe a condition in you. He also caps concurrent attention: his own best results come with two or three positions on, and he says you cannot honestly call 15 open positions high probability trades. Trade fewer markets and know them. One good trade a day was the first advice he got on the floor.
PRICE THE REWARD BEFORE YOU PRICE THE ENTRY
Defense first. Find where the stop technically belongs, which sets the risk. Then estimate the target using measured congestion, prior waves, Fibonacci levels, how much room the oscillator has left, or resistance on a higher time frame. Minimum acceptable payoff is 2:1 or 3:1 on day trades and at least 5:1 on longer holds; the book's sample money management plan sets a flat 3:1 floor. A high probability setup gets your maximum table size, a medium one gets medium size, and if the proper stop is too far away you either trade smaller or take zero.
PUT A CLOCK ON DEAD TRADES
Good trades work almost immediately. If a day trade has done nothing in 30 minutes, Link says you already know you are wrong; by 45 minutes he is out regardless of whether the stop was hit. This is aimed at the specific failure where a position down 10 cents gets ignored, becomes 40 cents, then a dollar, and by then you refuse to sell. On the profit side he takes money inside 80 percent of the market's average daily range rather than holding for the extreme, and he never places stops at obvious levels (round numbers, exact trendlines, prior highs) because those are where everyone else's stops sit.
TEST BEFORE YOU TRUST, THEN REVIEW NIGHTLY
His backtest gate: at least 30 trades in the sample or the result is chance, a profit factor of 1.5 minimum (over 2 is good), largest winner to largest loser of at least 1.5:1 and preferably 2:1 or 3:1, a drawdown you could afford to suffer twice over, and commissions plus slippage loaded in. In his own worked example, adding $15 round-turn commissions and $100 of slippage turned a system showing $7,025 profit into a $14,000 loss. Then review every night: open positions first, then the small losses you exited correctly (he reviews those before the winners, deliberately, to reinforce the behavior), then the mistakes, then the wins.
What it looks like Monday morning
Open a spreadsheet before you open a chart. Write down total capital, halve it, and take 5 percent of that half. That single dollar figure is now your maximum loss on any position. Then list the five to ten things you actually trade, pull the 14-day ATR for each, divide your risk number by the ATR, and you have a hard share and contract ceiling per name. Tape it somewhere. Link's point is that this takes about 20 minutes once and removes the worst decision you make under pressure, which is how big.
Then add two habits. Every night, write the specific names and levels you want tomorrow along with where you would be wrong, so the morning is execution instead of hunting. And put a timer on every intraday position: if it has not started working in 45 minutes, close it, win, lose, or flat. Those two changes plus the size ceiling are the entire practical yield of the book, and none of them require a new indicator.
| Market | 14-day ATR | Max contracts or shares |
|---|---|---|
| S&P 500 | $4,800.00 | 0 |
| Mini S&P | $1,000.00 | 1 |
| U.S. T-bonds | $1,100.00 | 1 |
| Crude oil | $750.00 | 1 |
| Corn | $175.00 | 7 |
| Gold | $300.00 | 4 |
| MSFT | $2.00 | 600 |
| IBM | $3.50 | 300 |
| DELL | $1.50 | 800 |
Risk per trade divided by the dollar value of the market's average true range; these are 2002 ranges and the table only works if you rebuild it with current ATR.
Zero is an acceptable choice as the number of contracts to trade.Marcel Link, High Probability Trading
Where it fails
- The book's own sample plan breaks the book's own rules. Chapter 15 says a daily cutoff of 2 to 5 percent of at-risk equity is a big loss. Twenty pages later, the worked Money Management Plan Ideas sets the daily stop at $1,500 on $15,000, which is 10 percent, and labels it as such without comment. The same sample plan sets risk per trade at 5 percent while the prose has just argued 2 percent is the professional standard, and caps total open risk at 20 percent while the surrounding text says half is the ceiling and under 30 percent is better. These are guidelines that argue with each other, not a system. You have to pick one set and be consistent yourself.
- The math prices out the reader most likely to buy it. Run his own Table 15-1 honestly. At $50,000 total capital and 5 percent risk on the $25,000 at-risk half, the maximum allowable size is zero contracts of the S&P 500, zero of the NASDAQ 100, and exactly one of most other futures. He states the minimum viable account as $25,000 to $50,000 for futures and $100,000 for equities. Anyone under that is told, correctly but unhelpfully, to trade less and expect less. The discipline framework is real, but the sizing engine at its center does not produce a tradeable answer for a small account, and the book never solves that.
- Every number is anchored to 2000 to 2002 volatility. The position table assumes AMAT with a $2.50 daily range, MSFT at $2.00, IBM at $3.50, crude oil ATR at $750. The chapter on setting goals uses AMAT with a $4 daily range as the reference example and mentions ARBA going from a $15 daily range to 50 cents as a curiosity rather than a warning. He also spends a full chapter arguing that cheap online commissions, real-time quotes, and 15 to 30 point intraday S&P swings are the new retail edge. Commissions went to zero, spreads compressed, and the edge went with them. The method survives; the calibration does not, and you must rebuild the ATR table from current data.
- He never tests his own setups to the standard he demands. Chapter 13 insists you never trade a system with fewer than 30 sample trades, a profit factor under 1.5, or untested slippage. Yet the 30-minute breakout system, the moving-average and trendline systems, and the multiple-time-frame confirmation routine are all published as entry and exit rules with no results attached. The crude oil worked example in Chapter 10 is a single hand-picked winner, and he says outright that the indicators he chose for it were random. There is no expectancy figure anywhere for high probability trading itself. The title makes a statistical claim the book never quantifies.
- The useful part is about 90 pages of 400. Link warns in the preface that he repeats himself on purpose. He does. Every one of the 18 chapters ends with a Becoming a Better Trader recap, a numbered do list, a numbered do-not list, and a Helpful Questions block, much of it restating the previous chapter. Chapters 5 through 9 are ordinary textbook technical analysis (moving averages, RSI, MACD, ADX, retracements, breakout patterns) that you can get free anywhere and that Link himself says is not where traders fail. The actual contribution is Chapter 10 (what qualifies as a trade), 11 (trading plan and game plan), 13 (backtest evaluation), and 15 (risk parameters). Buy it for those and skim the rest.
- It quietly does not work for part-time traders. The method assumes a nightly market review, a full day at the screen, two or three positions you can actively watch, a 45-minute intraday time stop, scaling in after 30 minutes, and rewriting the game plan at lunch. None of that is executable with a job. He never addresses it. If you cannot be at the screen during the session, the risk parameter chapters still apply but the trade management, which is half the book's value, does not.
- One printing, 2003, never revised. Every threshold on this page comes from the only edition McGraw-Hill ever published. Nothing here was updated for decimalization aftermath, the pattern day trader rule (in force since 2001 and never mentioned), electronic execution, or zero-commission brokers. The tooling advice is a period piece: buy TradeStation for $3,000, keep CNBC on all day, order paper charts from CRB and update them by hand, and use a live broker until you are experienced. Treat the numbers as Link's 2003 calibration, not as current settings.
Who it is for
Buy it if
You already have setups you believe in and a P&L that says otherwise, and you suspect the leak is sizing, overtrading, or the absence of anything written down. It lands hardest for a funded discretionary trader with $25,000 or more who has never built a position size table or a nightly game plan. Chapters 10, 11, 13 and 15 will pay for the book several times over.
Skip it if
You are hunting for an entry method or a proprietary edge, because there is not one here and Link says so. Skip it if you are trading a small account, since the sizing math returns unusable answers below roughly $50,000 and he offers no workaround. Skip it if you already keep a written trading plan, size by fixed fraction, and journal your trades, because that is the book's whole payload and you have it.
As an Amazon Associate I earn from qualifying purchases. It costs you nothing extra, and it does not change what gets recommended or what the verdict says.