OSRS Margin and ROI Explained

Two numbers get quoted constantly in flipping discussions, and they measure different things. Confusing them is the most common reason a trading strategy looks good on paper and does nothing in practice.

Margin: profit per unit

Margin is what you make on one item after everything is paid:

margin = (sell price − GE tax) − buy price

The tax matters. A raw spread of 1,000 GP on an item selling at 40,000 GP is only 200 GP of margin once the 2% comes off — 80% of the apparent profit disappears. Every margin figure on OSRS Helper is calculated after tax for exactly this reason.

Margin answers: how much do I make each time this trades?

ROI: profit per GP committed

Return on investment expresses that same profit as a percentage of what it cost you:

ROI = margin ÷ buy price × 100

ROI answers: how hard is my money working?

Why they disagree

Consider two items:

Item B has a margin ten times larger. Item A makes your capital work sixteen times harder.

With 8,000,000 GP you could buy 1,600 of Item A for 640,000 GP of profit, or 10 of Item B for 40,000 GP. The item with the smaller margin wins decisively.

This is why ranking flips by margin alone produces bad decisions. Margin without context tells you almost nothing.

Why ROI alone is also wrong

Now add buy limits, and the picture changes again.

Suppose Item A has a buy limit of 100. You can only buy 100 per four hours, so your 8% ROI applies to 500,000 GP of capital, not 8,000,000. Your actual profit per cycle is 40,000 GP — identical to Item B, despite the sixteen-fold ROI advantage.

The number that reconciles all of this is profit per buy limit:

profit per cycle = margin × buy limit

That is what our highest margin screen sorts by, and it is the closest single figure to "what is this item worth to me in the next four hours".

Which should you optimise for?

It depends on what is scarce.

If capital is your constraint — you have a small bank and more good items than money — optimise for ROI. Getting the most out of every GP is what matters, and high-ROI items recycle your money faster.

If slots and attention are your constraint — you have plenty of GP and only eight offer slots — optimise for profit per limit. You want each slot working on the item that returns the most per cycle, regardless of percentage.

Most players move from the first situation to the second as their bank grows. A trader with 2M is looking for percentage returns; a trader with 500M is looking for items that can absorb capital at all.

The number nobody quotes: fill rate

There is a third dimension that neither margin nor ROI captures, and it decides whether either of them is real.

An item with a 10% ROI that fills once a day is worse than an item with a 2% ROI that fills six times. The theoretical profit is a rate per completed cycle, and cycles that never complete produce nothing.

This is why volume is the first thing to check on any candidate. A margin you cannot realise is not a margin. Our item pages show 24-hour volume and a liquidity rating alongside every price, and the item page for anything you are considering will give you an estimate of how long one buy limit is likely to take to fill.

Putting it together

A good flipping candidate needs all four:

1. Positive margin after tax — the spread survives the 2%
2. Reasonable ROI — the capital is working, not parked
3. A buy limit worth having — the cycle profit justifies the slot
4. Volume to fill — both sides complete in a sensible time

Items that clear all four are less common than the margin lists suggest. Finding them is the actual skill in flipping, and it is what our screens filter for — every ranked market view on this site applies a volume floor before it sorts by anything else.

To check a specific trade, the GE profit calculator will give you margin, ROI, capital required and break-even in one go.

Screening one item, end to end

The theory above is only useful as a procedure. Here is the whole thing applied to a single candidate, in the order the numbers should be checked.

1. Volume first, before you look at price

   Open the item and read its 24-hour volume. Below roughly 3,000 units a day, stop — whatever the margin says, you are unlikely to complete a cycle when you want to. This check is first because it is the one that invalidates everything else.

2. Margin after tax, not the raw spread

   Take the insta-buy price minus the insta-sell price, then remove 2% of the sell price. On an item selling at 400,000 gp a 9,000 gp spread becomes 1,000 gp of margin. Always work from the after-tax figure.

3. Profit per cycle

   Multiply the after-tax margin by the buy limit. This is what the item is worth to you in the next four hours, and it is the only figure that lets you compare a cheap high-limit item against an expensive low-limit one.

4. Capital required

   Multiply the buy price by the buy limit. If that number is more than about a third of your bank, the item is too big for you — one stuck position should never freeze most of your money.

5. Sanity-check the chart

   Look at the last week. A margin that only appeared today usually means the price is moving, not that you have found something. A margin that has been stable for days is a margin you can plan around.

A worked comparison

Three candidates, screened the same way. The figures are illustrative round numbers rather than live prices, but the shape of the answer is what matters.

Ranked by ROI the runes win. Ranked by margin the weapon wins. Ranked by what actually lands in your bank in four hours, the armour wins — and the weapon is last despite having the largest margin of the three.

Then look at the volume row. The weapon trades 900 times a day across every player in the game; you are competing for a share of that with your eight units. The runes trade 2.4 million times. That is the difference between a number on a screen and money.

Where these numbers go wrong

The one-line version

If you remember nothing else: ROI decides what to buy when money is scarce, profit per limit decides what to buy when slots are scarce, and volume decides whether either number means anything at all.