What not to do with what you see

The pane guides each carry their own warnings. This page collects the ones readers trip over most, in one place. Every entry names the mistake, says why it is wrong, and points at the pane that prints the honest version.

At a glance

• What this page is. A list of nine readings the desk does not support.
• Why it exists. Most of them come from reading a ranking as though it were a price, or a count as though it were a judgement.
• What it is not. It is not a list of things to do instead. The desk's own view sits on the Read-out and in the Ask, not on this page.
• Key difference. A reading says what the desk found, and where it has a view, it states that too. The decision that follows stays yours.

The nine misreadings

• Reading a rank as a return. Every model figure on the desk is a place in a queue of about a thousand names. None of it is a price, a target or a percentage move. A name near the top of the queue was placed there against tonight's other names, and nothing more. Read Rank Not Return once and every ranking column makes sense.
• Treating the Bearish List as names to bet against. The Bearish List is a register of what not to hold. Choosing not to own a share can only cost you a gain you did not make. Betting against one can cost more than you put up, because a price can rise without limit. Use it to screen your own holdings, which is the job it was built for.
• Reading Down rate as a size of fall. On the Bearish List, a down rate near 51 in 100 says how often names read like this one finished down. It says nothing about how far a price falls, or whether it falls at all. Read it beside the Held column, which says how many trading days the signal has lasted.
• Reading the book as a list of what to buy. Trade Cards is the short list the firm published, with a side and a reason for each name. It places no order, connects to no account, and instructs nobody. What you do with it is your decision.
• Reading implied volatility as a forecast. It is what the market charges today for movement, quoted as a yearly percentage. Read it as an insurance quote, not a prediction. Volatility Weather draws the same price across every listed date.
• Reading the Market Map's colours as a ranking. On the Market Map, colour is the share price change over the period you picked, and tile size is what the company is worth. Neither is a ranking by Ohey, and neither is on the same scale as anything in the signals panes.
• Combining panes into one score. Nothing on the desk adds up. The ranking panes carry Ohey's model output and the options panes carry observed market prices. They come from two unconnected sources. The Ticker Study layout puts them side by side on purpose, and joins them in no other way.
• Reading the published example as tonight's book. A delayed book is labelled published example and carries the date of the trading day it came from. It is a real, complete book from a past trading day, kept as a record. Do not read its names as the ones struck for the next open. See Book History for the record of past trading days.
• Reading a long held count as safety. A signal held many trading days means the model kept finding the name in the same place, day after day. It says nothing about the share price, the company, or any risk you are taking. See Signal History.

Two habits that keep you out of all nine

• Ask what the number is a number of. A rank counts places in tonight's queue. Open interest counts contracts. A colour counts a price change. None of the three converts into either of the others.
• Read the pane's own guide before quoting a figure from it. Every pane guide defines its terms the first time they appear, and names the misreadings that pane attracts.

Frequently asked questions

Is any of this the desk telling me what not to trade? No. Each entry corrects a misreading of one figure. The desk's own view lives on the Read-out and in the Ask, labelled, not on this page.

A name is deep on the Bearish List and I own it. What now? That is a prompt to look again at your own reasoning, and nothing more. The pane never says sell, and it never says short.

Why does the desk refuse to give me one overall score? Because the two halves measure different things from different sources. One number over both would hide which half moved. The desk keeps both halves separate so you see which half moved.

Both time frames agreed on a name. Does that make it more likely to work? No. Agreement means the 45 and 20 trading day ranks landed on one side tonight. It is a fact about the readings, not a probability about the price.

Can the assistant tell me what to buy? No. Ask and it says so, then points you at the panes so you can read the figures yourself. See What this desk can and cannot tell you.

Related panes

• Rank Not Return, the primer behind the first three entries.
• Bearish List, the pane most often misread.
• Trade Cards, the signals and their labels.
• Reading The Options Suite, the rules every options pane follows.
• Market Map, where colour is a price change.