This is one of the few screens on the platform where the setup is fully described by four dropdowns, which means you can reproduce it exactly and argue with it precisely. That is worth more than a clever screen you cannot reconstruct. Below is what each control does, what each column means, and the one input the screen does not give you that the title of this piece depends on.
The four dropdowns, and the settings I would start from
The Squeeze view sits in the Insider Alpha navigation between Performance and Transfers. It describes itself as tickers where insiders are buying while short interest is elevated, pairing insider transactions with the bi monthly short interest report. Four controls sit above the results.
MIN SHORT PERCENT offered 5, 10, 15, 20 and 30 percent at capture. LOOKBACK offered 7, 30, 60 and 90 days. INSIDER DIRECTION was set to Buys. MIN TRADE VALUE offered Any, 25 thousand, 100 thousand, 500 thousand and one million dollars. There is a Refresh button on the right and a candidate count above the table.
Start at 20 percent minimum short interest, a 30 day lookback, direction Buys, and a minimum trade value of 100 thousand dollars. The reasoning behind each of those is worth stating rather than accepting. Twenty percent of float short is genuinely crowded rather than merely elevated. Thirty days keeps the purchase reasonably close in time to the short interest snapshot it is being joined against. Buys because a sale in a shorted name is a completely different situation. And a hundred thousand dollars because below that you are mostly picking up small ceremonial purchases that are common at every company and mean very little.
The columns, and which one is doing the work
The results table carried eight columns at capture: Ticker, Company, Insider, Filed, Trade value, SI percent of float sorted descending, DTC and Signal.
SI percent of float is the crowding measure and it is the sort key, so the top of the table is always the most heavily shorted name rather than the best setup. Do not confuse position on the list with quality. DTC is days to cover, which is shares short divided by average daily volume, and it answers a different question: not how crowded the short is but how long it would take the shorts to get out at normal volume. A name can be 25 percent short with a low days to cover because it trades heavily, and that is a far less explosive configuration than 15 percent short with a high days to cover in a name that barely trades.
Filed and Trade value together are the insider side. Filed is when the disclosure appeared, not when the purchase happened, and those are different dates. Trade value is the dollar size of the transaction, which is the number your minimum trade value filter acted on.
Then read the line under the title, which is the most important thing on the page and the easiest to skip. At capture it read FINRA short interest as of 2026-03-30, with 26,353 tickers covered. That is the date of the short data. The footer explains why it works this way: short interest is published twice per month, on the fifteenth and the last business day. Your short interest number is a fortnightly snapshot with a reporting lag, and the insider filing beside it may be considerably more recent. Always check that date before you treat a row as current.
Rising borrow is the half of this the screen cannot give you
The title of this piece names two conditions and the screen delivers one of them. There is no borrow fee column, no utilisation and no rebate rate in that table. I looked and it is not there, so I am not going to tell you the module supplies it.
Get it from your broker instead, and specifically from the same place you would go to borrow the stock. Most brokers that permit shorting will show an indicative annual borrow rate on the ticker, and some show whether shares are easy to borrow or hard to borrow. That is the number that tells you whether lendable supply is tightening right now, which is the thing published short interest fundamentally cannot tell you because it is a snapshot from up to a few weeks ago.
The practical routine is two windows. The screen produces a shortlist, and for each ticker on it you check the borrow rate on your broker before it becomes a candidate. A name where the fee has gone from low single digits to something eye watering has a live scarcity story. A name with 22 percent short interest and a borrow rate of a couple of percent is one where lenders are plentiful and nothing is tightening, which means the squeeze part of the squeeze thesis is not present regardless of how alarming the short interest looks.
Record the rate you saw and the date, because your entire thesis rests on the direction it moves, and you cannot know the direction if you only ever look once.
What a genuine open market purchase adds to a crowded short
Set the direction filter to Buys and you have still not guaranteed the transaction means what you think. The module's own coverage description lists what it ingests, and the categories are worth memorising: open market buys, open market sells, option exercises, and 10b5-1 plan sales. Those are not the same event and only one of them is a person choosing to spend their own money at today's price.
An option exercise is an employee converting a grant, often at a strike set years ago, and the shares are frequently sold the same day. It shows up as an acquisition and it tells you almost nothing about their view of the current price. A purchase under a 10b5-1 plan was scheduled in advance, which is the whole point of such a plan, so it also carries less information about what the insider thinks today. An unscheduled open market purchase at the prevailing price is the one that carries a view.
In a crowded short specifically, that distinction matters more than usual, because the population of insiders at heavily shorted companies contains a lot of people whose compensation makes them mechanically long and whose filings will therefore be full of grants and exercises. Open the actual filing from the row and confirm the transaction type before the position exists. The one minute this costs is the highest return minute in the whole workflow.
What the genuine purchase adds is not a prediction. It is a change to the shape of the risk. Shorts in a crowded name are exposed to a supply shock, and an insider buying and holding removes shares from the pool that can be borrowed against, since shares held by affiliates are generally not sitting in a lending program. That is a small effect in a large company and a real one in a small float. It is also a signal to the other shorts, which is why the reaction is sometimes larger than the arithmetic justifies.
The empty screen is normal, and the button that gets people into trouble
At capture, with 15 percent, 30 days, Buys and 100 thousand dollars set, the screen returned zero candidates and displayed a message saying there were no squeeze candidates in this window, suggesting you lower the minimum short percentage or widen the lookback.
That suggestion is the trap. It is helpful interface design and it is terrible position selection, because loosening a filter until something appears converts a rare setup into a common one and then trades it as if it were still rare. If you go from 15 percent to 5 percent, you are no longer looking at crowded shorts. If you go from 30 days to 90 days, you are pairing insider purchases with short interest data that may predate them by months.
Do the opposite. Keep the filters and check the screen twice a month, ideally just after each short interest publication. Note down the days it returns nothing, because that count is the honest base rate for this setup, and it is the thing that tells you the setup is worth acting on when it finally appears. A screen that always has candidates is a screen that has been tuned to produce them.