In one sentence
Weight of money compares how much unmatched cash sits on each side of the ladder, and VWAP (volume-weighted average price) tells you the average odds at which money has actually been matched.
How it works
Weight of money is the oldest trader's gauge on Betfair. You add up everything in the blue "available to back" column and everything in the pink "available to lay" column, and see which is heavier. Many traders read a heavy pink side (lots of backers queuing) as support for a price to shorten, but the reading is a rough one.
VWAP looks at what has been matched rather than what is waiting. Each traded price is weighted by the amount matched there, so a £2,000 trade counts four times as much as a £500 trade. It gives a fairer "average price" than just looking at the last traded price.
The two work well together. Waiting money can be pulled in a second, but matched money is real, so VWAP acts as a reality check on a weight-of-money reading.
The maths
- L: total money waiting in the pink "available to lay" column.
- B: total money waiting in the blue "available to back" column.
- WoM: the share of waiting money on the pink side, between 0 and 1.
- pi: the odds of the i-th matched trade (or price level on the traded-volume chart).
- vi: the amount matched at that price.
In plain English: weight of money is a ratio of queued cash; VWAP is the average matched price, weighting big trades more.
Worked betting example
In the last ten minutes before kick-off, a football team's Match Odds price traded:
- 3.00 for £2,000
- 3.05 for £1,500
- 3.10 for £500
- 2.98 for £1,000
Total matched = £5,000.
Price times volume: 6,000 + 4,575 + 1,550 + 2,980 = 15,105.
VWAP = 15,105 ÷ 5,000 = 3.021.
Averaging in probability terms instead (1 ÷ odds, weighted by volume) gives 33.11%, which converts back to 3.0205. With prices this close together the two methods agree, but across a wide range, such as a drifter from 4.0 to 8.0, they differ, and the probability version is the one that behaves properly.
Now the ladder shows £9,000 waiting on the pink side and £6,500 on the blue side. WoM = 9,000 ÷ 15,500 ≈ 0.58. The current best back price is 3.05, above the 3.021 VWAP, so the price has drifted relative to where most money was matched even though the pink side looks heavier. That kind of disagreement is a reason for caution, not a trade signal.
Where it's good
- A quick sense check on whether a price move is backed by matched money or just shuffling of unmatched orders.
- Judging your own fills: did you get on better or worse than the market's VWAP?
- Big-league football Match Odds in the hour before kick-off, when matched volume builds fastest.
- Building features for short-term trading models.
Limitations and pitfalls
- Weight of money is easily faked: large offers placed well away from the price and pulled later cost nothing.
- Much of the money deep in the ladder will never be matched, so totalling the whole ladder overweights noise.
- VWAP looks backwards; it tells you where the market was, not where it is going.
- Averaging odds directly is biased when prices vary widely; average implied probabilities instead.
- Neither measure tells you anything about the true chance of the outcome.
- Over a short window a single large trade can dominate VWAP, so choose the window with care.
How to build it
- Betfair's API returns traded volume by price and the available-to-back and available-to-lay ladders; betfairlightweight in Python parses both.
- pandas makes rolling VWAP windows simple (group by time window, then sum price times volume over sum volume).
- Tip: compute VWAP on implied probabilities, and restrict weight of money to the top three to five levels.
Related methods
- Order book imbalance is the sharper, top-of-book version of weight of money.
- VPIN uses traded volume to detect one-sided, possibly informed flow.
- Closing line value compares your price with the final market price, much as VWAP compares it with the average.
- Moving averages and smoothing applies the same averaging idea over time.