This article considers how to price the index as a whole and how to value an IPD trade. It’s more maths heavy than the last one so get your calculator out.
Like the first article — These are just my opinions and this could all be wrong! I welcome anyone who thinks so to comment on it. If you spot any errors let me know I haven’t had much time to double check it. Also if you aren’t familiar with the term dumb money please read my first article. It’s not meant to offend anyone!
FI Top Trumps
The price you see from FI on an IPO is basically a bet by them that they will pay-out less than this on that future over the futures lifetime. Anyone is free to take the bet from them.
FI’s profit from a future(P) will be the futures price at issuance(I) + trading fees & spread achieved on that share throughout its life time(T) — dividends over the lifetime of the future(D). You could break T down to two different parameters one for the fees achieved from the spread and one from the fees achieved from the 2% fee. You could also break D down into different parameters — one for each of the dividend types. For this article though we will stick with ;
(This doesn’t consider the fact that FI could be using the money they collect from issuing futures to generate more revenue in activities external to the index. I’m ignoring these revenues as we have no idea what they are)
From FI’s perspective they see each future in each player in terms of P I D and T. The image above is a theoretical Messi future issued at £1. Right of the image is the same future after 10p in dividends and a single transfer at 2% (To keep it simple I’m assuming no difference in buy and sell and price so FI wont profit or lose from the spread in this scenario. If there was a spread this would be accounted for in the T value). If this was the only trade that took place over the life of that future (so basically the only trade that took place before the player retired) then FI would have made 92p in profit by issuing it.
If you wanted to know FI’s current P value on Messi as a whole, you would add up the figures from every Messi future issued at any price. If you wanted to know the P value for the market you would do this for every future in every player. Basically, the whole market can be described by P= I+T -D. Also, there is enough information out there for us to model every one of these parameters to date except T. We can’t model T because we don’t know how many non-instant sell trades take place in a day. Instant buy or sell trades move the price (A 900 future trade moves the price 1p ) so we can approximate the 2% fee on these trades as we can measure price movements. To my knowledge Non — Instant sell trades don’t move the price so we have no way of knowing how many take place and can’t model the fees collected. If I am wrong on this and someone out there has been able to track the non-instant sell volumes, then we can input it into the model and we should be able to get the P value for the whole index in it’s present state.
What observations can we make from the model?
Basically, they lose money when I+T is smaller than D. To avoid this, they must be able to model D and T in advance, so they know what to make I in order to achieve a positive P. What this tells us is that they know with a degree of certainty what the futures’ dividend pay-outs will be over the futures’ lifetime and the fees that future will generate. They have to know this (to some degree) on the day they issue the future. Originally I thought that they had a way to model D using historical data sets but now I am not so certain.
T probably allows them to get away with being less accurate in modelling the D value. It’s a reasonable assumption to make as we have been told that the average share is only held for 2 days, so across the whole index the combined T value must be huge! Large T means you can be wildly inaccurate in the D value and still end up with a positive P.
Note that a sharp punter who spots value in a future on IPO day and holds with the idea of never selling would then pose a great risk to FI as this type of player results in a small T value as the player will only roll the share over once every three years. FI better hope that D was accurate when they set the value for I and that they priced for T only increasing once every three years or else after the end of the players career P will be negative.
The safest way to protect against negative P is to increase the IPO price (I). This actually results not only in an increase in I but possibly in an increased trading fee value (T). The reason being if dumb money doesn’t understand they are being deprived of value with the price hike then they will keep trading the new IPOs in similar volumes to how they traded old IPOs. This means the value of T increases by the same proportion as the price hike.
Even if they don’t buy in the same volume as before FI only need a fraction of the volume to achieve the same T value. For example if FI IPO’d a player at £1 and two people bought then sold him at £1 they would only make 4p (2%fee x £1 x 2 sell trades). If FI can IPO players at £2 then one person buying and then selling at £2 gives them 4p. The number of traders purchasing the share has halved but the fees FI receive remain constant. Take this scenario further and imagine the person who didn’t buy after the price hike chose not to do so because they knew that over the lifetime of the share it would return less than what they paid for it. What FI have just done is to remove sharp punters from the market, sharp punters that might cost FI money in the long run. But they haven’t lost out in fees.
Hikes in IPO prices without an equal increase in dividends ultimately results in a lower expected value over the futures lifetime, it probably makes them negative for a lot of futures currently being IPO’d but I can’t tell for certain. It’s possible that FI anticipate another MB and PB dividend hike down the line and as such they are issuing higher IPOs now. When the hike occurs then value will return to a lot of the valueless futures on the market.
From a mathematical standpoint the ideal market for FI would consist of dumb money consistently speculating on futures based on what happens in the world of football without ever understanding the underlying mathematics that give the futures their value. This market would be ruled by “Relative Value” dumb money traders taking obscene IPO prices that have massively negative expected values and hoping that whenever an event occurs people trade it meaninglessly without realising that it hasn’t affected the players total pay-out prospects thus increasing the T value on the shares.
Note that IPD’s cause a lot of trading so massively increase T. However, FI aren’t paying out much on them so D barely moves, and P is massively increased.
Why I don’t like IPD’s
When the IPD hike happened alongside the share split I asked someone on the forum to point to a player that had a positive intrinsic value based on IPD’s alone. I was told to go look at Fabio Quagliarella. So I have!
Figure 1–30 Day IPD pay-out. The pay-out in pence that you would have received over 30 days from the date (x-axis) had you purchased Quag futures on that date. As Quag is a striker a pay-out event is a goal or an assist and (at the time of me writing this) is worth 1p.
Quags peak was 7 pay-out events. You would have made 7p from him had you purchased him on the 30/11/2018. You would have received IPD’s up until and including 30/12/2019. He played 6 matches over this time frame.
So does the bet have a positive intrinsic value? Does Quag return more in IPD pay-outs over a 30 day period than you pay in fees (spread + exit fee). The reason the spread is included is because for the bet to have a positive intrinsic value you shouldn’t have to rely on market volatility caused by other people moving the price to make the trade profitable. For there to be a positive intrinsic value to the trade the trade should be profitable without any market movement and the IPD’s should be greater than the trading fees. So Pay-out > Spread + Fees.
Trading the expected movement in an IPD player would be a perfectly legitimate way to trade but you are trading market volatility by trading like this and not intrinsic value. In theory volatility should only exist as market participants chase bets that have intrinsic value. If IPD’s don’t have a positive intrinsic value in and of themselves no one should be trading them and there should be no volatility to trade.
Trading fees — Exit fees are 2% except for a player below 50p where FI charge 1p. This means trading in players below 50p incurs a greater that 2% fee.
Spreads — For the scenarios below I have assumed a 2p spread. This a fair assumption (not to harsh not to generous) as most players up to about the £1.20 range have a 2–3p spread. Note again though that lower priced players are hurt by this spread being constant as it represents a greater percentage of their price. So you are being charged more to take bets in them
7 Payout Events
The above table shows the net yield relative to purchase price for any future given a fixed 2p spread if he scores 7 pay-outs over the 30 days. Its shows this net yield at different purchase prices. At 7 pay-out events you are still 3% profitable at £1.00
4 Pay-out events
At 4 payout events you would be break even at a purchase price of 80p. Note that Quags was the golden boot winner in Serie A last season and due to his age traded at sub £1. Most players getting 7 pay out events don’t trade below £1. If they do it won’t be for long! If you want to work out whether a player is good as an IPD hold or not then you have to look at their price relative to their potential to achieve N amount of payout events over the 30 day period.
For example lets assume that a player has a 50% chance of scoring 0, 30% chance of scoring 1, and a 20% of scoring 2 payout events. For simplicity assume that no other scores are achievable and remember 1 score event = 1p. You would calculate the expected value of this bet as follows
(0.5 x 0) + (0.3 x 1) + (0.2 x 2) = 0+0.3+0.4 = 0.7
This means the fair price is 0.7p . So if you can make the bet for less that 0.7p its worth making as you are taking value. You will still loose half of the time as 50% of the time the player doesn’t score. Continue to do this over a series of bets and you will finish up in the long run and FI will finish down. IPD’s are interesting because the cost of the bet is actually paid out upon selling the shares not upon their purchase. The cost of the bet is the trading fee + spread. So in this case you would want trading fee + spread to be less than 0.7p. In a real situation there is of course more than 3 score possibilities — you have to decide what the probability of a player achieving each score is. For an outfield player you can do this by looking back at the historic goals and assists they get over a 30 day time frame. For a keeper you would be looking at stats around goals conceded.
I’m doubtful many players have a positive intrinsic value from an IPD perspective and you will likely need the price to move in your favour to make a return on the bet.
Relating IPDs to P=I+T-D
You can make a 3 year bet on a player and only have to pay T (trading fees) once at the end of the three years. In that time, you are eligible for all PB and MB pay-outs. However, from an IPD you are only eligible for 30 days of payout before you must pay exactly the same in T in order to be eligible for the next 30 days of IPD’s. Its crazy, particularly at the upper end of the index where T is huge and the pay out from IPD is tiny. I have heard FI on their official podcast tipping players at the upper end of the index for IPD’s. It’s nonsense — remember Quags table at 7 payouts– At £1 purchase price you get 3% and he only pulled that off over one 30 day period. Note that at the time of writing Messi is circa £5. So he needs 5 pay-out events for 1% return and 10 pay-out events just to cover the 2% fee (assuming no spread!). Glancing at his chart on Transfermarkt last season I think 14 events was his biggest 30-day payout. You will want his price to move up over the same time frame to make that trade work most of the time.
Hope this was useful — Next time I’ll talk about why FI should allow us to short the market and the problems this would solve. Good luck trading.