Giancarlo Stanton’s Opt-Out Caps His Trade Value
With the Marlins looking to cut payroll, Giancarlo Stanton seems likely to be traded at some point this winter. Stanton not only takes up a lot of that payroll, but his talents are desirable to other teams. The grounds for a trade are obvious. At the same time, the 10 years and $295 million remaining on his contract present a roadblock to acquiring him. So does his no-trade clause, which he could use to block a trade to one of the clubs willing to take on his salary.
Another concern is Stanton’s opt-out clause. While less of an obstacle to the eventual completion of a deal, it’s a factor. By the terms of the opt-out, Stanton will have the choice, in three years, either to become a free agent, or exercise a player option for seven years and $218 million. In late June, when Stanton had a 131 wRC+ and looked to be on pace for “only” a four-win season, the opt-out didn’t seem to matter too much; the probability that he’d exercise it seemed pretty low. A monster second half and MVP Award later, though, and that opt-out is back in play, and it negatively affects Stanton’s value.
Without revisiting whether player opt-outs benefit a team (Dave Cameron wrote about it over the course of multiple posts two years ago with the last one here), let’s just dig into the numbers a bit and see what they say. Last week, Eno Sarris advocated for trading for Stanton, noting that if Stanton ages normally, there’s a big surplus on the contract. Sarris also noted that, even if Stanton ages poorly, the contract would remain pretty close to the current market rate.
Here’s an estimate of Stanton’s value were he to age normally, starting with $9 million a win. I’ve used Steamer’s win projection (5.3 WAR) for Stanton’s 2018 season.
| Year | Age | WAR | $/WAR | Est. Contract | Actual Contract |
|---|---|---|---|---|---|
| 2018 | 28 | 5.3 | $9.0 M | $47.7 M | $25.0 M |
| 2019 | 29 | 5.3 | $9.5 M | $50.1 M | $26.0 M |
| 2020 | 30 | 5.3 | $9.9 M | $52.6 M | $26.0 M |
| 2021 | 31 | 4.8 | $10.4 M | $50.0 M | $29.0 M |
| 2022 | 32 | 4.3 | $10.9 M | $47.0 M | $29.0 M |
| 2023 | 33 | 3.8 | $10.9 M | $41.6 M | $32.0 M |
| 2024 | 34 | 3.3 | $10.9 M | $36.1 M | $32.0 M |
| 2025 | 35 | 2.8 | $10.9 M | $30.6 M | $32.0 M |
| 2026 | 36 | 2.3 | $10.9 M | $25.2 M | $29.0 M |
| 2027 | 37 | 1.8 | $10.9 M | $19.7 M | $35.0 M |
| Totals | 39.0 | $400.6 M | $295.0 M |
Assumptions
Value: $9M/WAR with 5.0% inflation (for first 5 years)
Aging Curve: +0.25 WAR/yr (18-24), 0 WAR/yr (25-30),-0.5 WAR/yr (31-37),-0.75 WAR/yr (> 37)
As structured, the deal looks pretty good for the team, with a surplus of over $100 million. Except for one thing, though: this version of reality isn’t ever likely to occur. In three years, Stanton will be just one year older than J.D. Martinez is right now. If he puts up three five-win seasons in a row — better than what Martinez just did in one season — his value on the free-agent market is going to exceed the $218 million he is owed. In that case, Stanton would almost certainly opt out, cutting the surplus by 30%.
The above case represents just one scenario, though. To play this out, let’s consider a bunch more and see where it leads. For the purposes of this exercise, let’s assume the 5.3 WAR figure is a good estimate of Stanton’s present talent. Note that the higher estimate you put on his talents, the more likely Stanton is to opt out. If you put a six-win talent estimate on Stanton and give him 18 WAR over the next three seasons, that place among the greats. The only primary right fielders to do better than that from 28 to 30 since 1947 are Hank Aaron, Stan Musial, and Frank Robinson. Even in that scenario, the surplus is under $100 million. There’s a limit on how much high-end value Stanton can provide because of that opt-out.
To examine the consequences of the opt-out, we need to consider a wide range of outcomes. Here, I’ll consider scenarios in which Stanton (a) ages well, (b) ages normally, or (c) Stanton ages poorly. I’ll also consider the varying outcomes regarding Stanton’s current talent level, setting 5.3 WAR as the average outcome.
For the first exercise, we will consider Stanton’s value if he ages normally. Here are some possibilities of his current talent level and a rough estimate of the likelihood of those outcomes.
| Percentile Outcome | WAR | Percent Chance of Outcome |
|---|---|---|
| 90% | 8.8 | 4% |
| 80% | 7.6 | 8% |
| 70% | 6.6 | 12% |
| 60% | 5.8 | 16% |
| 50% | 5.3 | 20% |
| 40% | 4.8 | 16% |
| 30% | 4.1 | 12% |
| 20% | 3.1 | 8% |
| 10% | 1.9 | 4% |
We will use this same distribution throughout. As you can see, there’s a 52% chance that Stanton’s talent level is somewhere between 4.8 and 5.8 WAR, with the higher and lower estimates occurring less frequently. What we will now do is put all of those scenarios through the calculator, determine if Stanton would opt out, and find a surplus value that we will weight by the percentage chance of that outcome.
The first step looks like this.
| Percentile Outcome | WAR | Percent Chance of Outcome | Value Before Opt Out ($/M) | Salary Before Opt Out ($/M) | Salary After Opt Out ($/M) | Value After Opt Out ($/M) | Opt Out |
|---|---|---|---|---|---|---|---|
| 90% | 8.8 | 4% | 249.7 | 77 | 218 | 516.4 | Y |
| 80% | 7.6 | 8% | 215.6 | 77 | 218 | 425.2 | Y |
| 70% | 6.6 | 12% | 187.3 | 77 | 218 | 349 | Y |
| 60% | 5.8 | 16% | 164.6 | 77 | 218 | 288.2 | Y |
| 50% | 5.3 | 20% | 150.4 | 77 | 218 | 250.2 | Y |
| 40% | 4.8 | 16% | 136.2 | 77 | 218 | 212.2 | N |
| 30% | 4.1 | 12% | 116.3 | 77 | 218 | 159.0 | N |
| 20% | 3.1 | 8% | 88.0 | 77 | 218 | 87.7 | N |
| 10% | 1.9 | 4% | 53.9 | 77 | 218 | 30.8 | N |
If Stanton’s value after the opt-out is greater than his salary, then he opts out. In the above scenario, Stanton would opt out around 60% of the time.
Now let’s look at the surplus value that Stanton provides in these scenarios, beginning with a normal aging curve.
| Percentile Outcome | WAR | Percent Chance of Outcome | Surplus with Opt-Out Clause | Value of Surplus | Value of Surplus with no Opt-Out Clause |
|---|---|---|---|---|---|
| 90% | 8.8 | 4% | 172.7 | 6.908 | 18.844 |
| 80% | 7.6 | 8% | 138.6 | 13.816 | 27.664 |
| 70% | 6.6 | 12% | 110.3 | 13.236 | 28.956 |
| 60% | 5.8 | 16% | 87.6 | 14.016 | 25.248 |
| 50% | 5.3 | 20% | 73.4 | 14.68 | 21.12 |
| 40% | 4.8 | 16% | 53.4 | 8.544 | 8.544 |
| 30% | 4.1 | 12% | -19.7 | -2.364 | -2.364 |
| 20% | 3.1 | 8% | -119.3 | -9.544 | -9.544 |
| 10% | 1.9 | 4% | -210.3 | -8.412 | -8.412 |
| TOTAL | $50.9 M | $110.1 M |
In the best scenarios for Stanton, he opts out of the contract after three years, giving a team surplus value only for the first three years. That cuts the surplus from the $100-plus million like we saw in the first chart down to $50 million.
Now let’s repeat the process above — except, this time, we’ll assume that Stanton ages well. The opt-out situation does not change at all here, as he is presumed to hold steady from ages 28 to 30 before the opt-out anyway.
| Percentile Outcome | WAR | Percent Chance of Outcome | Surplus with Opt-Out Clause | Value of Surplus | Value of Surplus with no Opt-Out Clause |
|---|---|---|---|---|---|
| 90% | 8.8 | 4% | 172.7 | 6.908 | 21.9 |
| 80% | 7.6 | 8% | 138.6 | 13.816 | 33.776 |
| 70% | 6.6 | 12% | 110.3 | 13.236 | 38.136 |
| 60% | 5.8 | 16% | 87.6 | 14.016 | 37.472 |
| 50% | 5.3 | 20% | 73.4 | 14.68 | 36.4 |
| 40% | 4.8 | 16% | 129.8 | 20.768 | 20.768 |
| 30% | 4.1 | 12% | 56.7 | 6.804 | 6.804 |
| 20% | 3.1 | 8% | -47.7 | -3.816 | -3.816 |
| 10% | 1.9 | 4% | -173 | -6.92 | -6.92 |
| TOTAL | $79.5 M | $184.5 M |
If Stanton ages well, the surplus goes up to $80 million, but the opt-out costs the team more than $100 million because of the lost value.
Now let’s take a look at what happens if Stanton ages poorly. If that were to occur, he’d start to show his age a little bit at age 30 and the opt-out situation changes. He only opts out below at the 70th percentile outcome or higher, which happens 24% of the time.
| Percentile Outcome | WAR | Percent Chance of Outcome | Surplus with Opt-Out Clause | Value of Surplus | Value of Surplus with no Opt-Out Clause |
|---|---|---|---|---|---|
| 90% | 8.8 | 4% | 165.4 | 6.616 | 13.972 |
| 80% | 7.6 | 8% | 131.3 | 13.232 | 17.92 |
| 70% | 6.6 | 12% | 102.9 | 12.348 | 14.34 |
| 60% | 5.8 | 16% | 36 | 5.76 | 5.76 |
| 50% | 5.3 | 20% | -10.8 | -2.16 | -2.16 |
| 40% | 4.8 | 16% | -54.9 | -8.784 | -8.784 |
| 30% | 4.1 | 12% | -110.5 | -13.26 | -13.26 |
| 20% | 3.1 | 8% | -177.2 | -14.176 | -14.176 |
| 10% | 1.9 | 4% | -238.6 | -9.544 | -9.544 |
| TOTAL | -$10.0 M | $4.1 M |
We see that, if Stanton ages poorly, teams no longer get the value of the contract due to the opt-out. Without the opt-out clause, the contract still has surplus value, but the opt-out drops the value by around $15 million.
Now, we can’t just average the three scenarios to arrive at an estimated “cost” of the opt-out, because the likelihood of each scenario isn’t the same. We could come up with a number of different scenarios to find a value for the opt-out, but how about we assume that Stanton’s current contract is roughly market value and come up with a scenario to see how front offices might be viewing Stanton’s potential as he ages.
For these purposes, we will say Stanton ages well 5% of the time, ages normally 15% of the time, and ages poorly 80% of the time. To find your own valuation of Stanton, you can choose different figures. If you think Stanton is more likely to age normally, you are going to rate Stanton as a much more valuable player.
The scenario here is a rough approximation of his valuation if his current contract is market value.
| Contract Status | Surplus |
|---|---|
| With an Opt-Out Clause | $3.6 M |
| Without an Opt-out Clause | $29.1 M |
| Difference | $25.5 M |
If Giancarlo Stanton is a 5.3 WAR player right now and his contract is market value, that opt-out is worth around $25 million. Continuing with these assumptions, Stanton opts out 31% of the time, which seems a bit more reasonable than the 60% under the “ages well” and “ages normally” scenarios. If Stanton doesn’t opt-out, just 31% of the time does the team get value on the contract. As we might expect on a contract that has an expected value in line with the terms, it is pretty close to a 50/50 split on whether it works out for the team.
Most of the time, Stanton does not opt out, and in the vast majority of those circumstances, the team doesn’t get full value from Stanton. When Stanton does opt out, the team misses out on a good chunk of value. The better you believe Stanton is now and will be in the future, the more a team misses out with that opt-out clause. We often assume that the team with the highest valuation of a player will make the best offer, but in this case, a high valuation is going to put a cap on the value a team believes it will receive due to the opt-out clause. That clause limits the ceiling a team will receive from Stanton while retaining the risk that comes with a $295 million contract.
Craig Edwards can be found on twitter @craigjedwards.
The likelihood distribution on your current WAR projections seems pretty reasonable, but what’s the explanation for the aging assumptions? Why is there an 80% chance he ages poorly, and only a 5% chance he ages well?
I used those figures to fit the assumption that Stanton’s contract is currently at market value, which seems to be pretty close to the prevailing view. We could do a lot more to see how realistic that is and come up with a bunch of comps for Stanton, but that is probably a whole other article.
Fair enough. But teams’ valuation of him could also mean they have a lower estimate of his current talent projection.
In any case, I love that you took the time to do this kind of analysis. It’s what we should think about in terms of any transactional analysis. Not just a chart with a year-by-year WAR projection (which we’ve seen on Fangraphs a thousand times), but also considering the likelihood of a range of different projections and different aging curves.
I like how you use the idea his contract is close to market value to assign probabilities.
As Yirmiyahu said, it might mean that teams have a lower estimate of Stanton’s current talent.
However, 5.3 WAR seems reasonable enough to me.
I strongly suspect that it is $/WAR figure that is off.
If $9M/WAR is the true current market price, then there is no reason that the Reds would not offer Cozart QO because Cozart would never take it.
You’re right, $9M per WAR is suspect. It’s more like $11M.
https://www.fangraphs.com/blogs/the-recent-history-of-free-agent-pricing/
Also, you should pretty much never use decisions made by the Reds front office as evidence of how the baseball market as a whole operates.
So according to this figure,
Mets expected 10.5 WAR from Cespedes for four years
Cards expected 7.9 WAR from Fowler for five years
Rockies expected 6.7 WAR from Desmond for five years
Dodgers expected 6.1 WAR from Turner for four years
Indians expected 5.7 WAR from Encarnacion for three years
Astros expected 5.0 WAR from Reddick for four years.
I will take over on all of them.
There is a massive difference between what the teams have been paying per WAR in the past and what the team is willing to pay per WAR in the future.
As great as Trout is, I doubt any team would go remotely close to paying $78M AAV for current Trout (using 8.7 steamer projection and $9M /war) let alone $98M AAV using $11M.
That is a bunch of words that does nothing to invalidate the results of the study I linked to, which is based on what teams actually paid, the actual performance of the players, and the average annual inflation of the ratio between the two. You could learn a lot if you actually read it.
Trout’s contract wouldn’t approach a $78M AAV because he would be given a long-term contract that he would be expected to decline during. He would get something like a $40M AAV which would be a surplus at the beginning of the contract and a deficit on the back end. He’d probably also have an opt-out or three, which is extra value that doesn’t show up in the AAV.
The reason you’re so high on the WAR estimates for the players above is also because you’re failing to factor in expected decline (except for Turner, that was just a steal). It’s also because I said it was around $11M this year, last year it was more like $10M (and again, that’s simply a fact based on the production of all players who started the year with over 6 years service time, as well as all of their respective salaries).
Clarification
1. I was using $10M for last year’s analysis
2. I am not saying ‘I’ expected more than 10.5 war at the signing. I am saying that the Mets were surely higher on Cespedes than 10.5 war when they signed him.
Using Trout’s next year projection of 8.7 war and a harsh decline of -0.5 war every year starting next year when he is only 26, do you know how long a contract has to be to get the AAV all the way down to 40M if we were to use $9M/war and inflation of 5% for the first 5 years and zero afterwards?
20 years.
It covers 2018 (age 26) to 2037 (age 45) where Trout is expected to produce 79 War and the estimated contract price is $823M for 20 years with AAV 41M.
Do you think this comes remotely close to what Trout would receive if he were to be a FA this winter?
No, I don’t think he would get 20 years. He’d just get a much higher AAV. This still does nothing to refute the objective facts spelled out in the link I provided. You can argue until you’re blue in the face, the math doesn’t care.
bmarkham, I think tung twista is on to someting here. I’ve been thinking about this in the context of the Eric Hosmer contract thing and I think the $10 million/WAR figure is being thrown around too easily. This doesn’t mean that Swartz’s findings are wrong (they are likely correct, technically speaking) but that we do not interpret the findings correctly when discussing that figure.
1) Part of the problem is that there are some really influential outliers in the history of free agent pricing. Swartz actually mentions this in one of the later articles in this series, how there are deals signed 10 years ago that no longer look bad because deals for Pujols, Hamilton, etc have reset the actual value of $/WAR. This should be a huge red flag. In these analyses, Jason Heyward and Robinson Cano are part of the $/WAR figure for several years because they are 3 year averages.
This, by the way, does not actually refute any of Swartz’s analysis. It just means that the expectation of what a team is going to get is not likely what this figure is capturing.
2) Swartz also notes that this is not a classic supply-and-demand labor market. The amount of money teams spend appears to be driven by league revenue, while player performance appears to be driven by really good players signing extensions early and no longer being part of the market. So there is a scarcity of talent, and a rising price tag, but the two do not appear to be related to each other.
This whole thing suggests that tung twista is right to be skeptical of the $/WAR figure. Swartz is probably right, but we are probably wrong for how we interpret his work. Many teams are probably not interpreting $10 million/WAR as a good deal, and neither should we. In fact, the results of the study suggest that finding good value in the free agent market is harder than ever, since teams are throwing more money at an increasingly shrinking talent pool. In the context of spending your money elsewhere (on extensions, for example) it looks like the argument isn’t that teams should be spend more money per win on the free agent market, but that teams should spend less overall in the FA market and more elsewhere.
The better way to analyze $/WAR is that salaries are determined through a process of normative comparison. This is something sociologists have started looking into with CEO pay. Here, CEO salaries appear to be totally disconnected from any objective real-world information about “value” and instead are determined by comparing salaries to previous salaries. And that is what we see at the top of the free agent market here too, right? Agents and teams negotiate based on benchmarks about what other contracts have received (if Jordan Zimmerman got this, then this is what Jake Arrieta deserves).
so yeah, bmarkham is right that Swartz’s math doesn’t lie, but tung twista is absolutely right to say something is off about how we’re using that figure.
The value of a point of WAR is not predictive, it is based on the previous season’s production and subsequent AAV of the contracts signed by the free agents.
The outliers don’t establish the value of a point of WAR. The cumulative production and cumulative AAV dollar value of the contracts signed established the value of a point of WAR.
A free agent player that produced a negative WAR the previous season does not owe his new team money. The free agent that produced a 7.6 WAR will not sign for $68.4 million just because a point of WAR had a $9 million value for 2017.
What a point of WAR does do is establish the surplus value of a player’s contract. In the article above, the “market value” of the player’s contract is based on what a point of WAR is worth in each season of his contract vs how much he is paid that season of his contract.
If Stanton puts up a 5.3 WAR for the next 3 seasons, a figure suggested above, and is paid $77 million for those 3 seasons, then Stanton has produced $143.1 million in value based on $9 million per WAR. He has produced a market value of nearly double his salary paid. Since the value of a point of WAR is subject to salary inflation, it is realistic to say that he would have produced a market value of more than double his salary.
If you go by the formula here and are NOT generous with when he turns 31 I still end up with 106 WAR instead of your 79 WAR. And a total contract of 1.13 Billion averaging 56.5 million a year. It would take until he is 53 before he would average 40 million a year.
Trout breaks the formula by a very wide margin.
Good comment, but I strongly suspect that the big difference is in assumptions on aging curve / injury risk.
I’d be curious if anyone has ever calculated aging curves specific to players who are already at a high level (something like at or above 4-5 WAR). By definition, that excludes some players who had an unexpected leap in production in their late 20’s – e.g, Justin Turner or J.D. Martinez – who are part of the average aging curve. There’s also the basic definitional math that a player who misses time due to injury compiles zero WAR during his absence whether he’s a 5 WAR player or a 2 WAR player. Since I’m pretty sure that part of the aging curve is that older players on average miss more time due to injuries, that component of the aging curve ought to be different for players with different WAR forecasts when healthy.
In the case of Stanton specifically, I think that teams must have concerns about a player with his history of missed time. A couple of those absences were due to hit by pitches, specifically the last 15 or so games of 2014 and a bit over half of the 2015 season. That still leaves, however, playing 123 games of less in each of 2012, 2013, and 2016 due to various other injuries (hamstring, knee, groin, side, shoulder, foot). I think that the lower body injuries in particular are something of a red flag for a team projecting the future health of a 6’6″, 245 pound slugger in his late 20’s.
Wouldn’t a team actually prefer him opting out? First it probably means he was pretty good those 3 years if he choses to opt out, and for surplus value it probably is pretty good too because you might lose some surplus value in year 4 and 5 but you are off the hook for the negative value he likely produces in years 6-10.
Now it obviously isn’t ideal the the opt out is one sided so the risk is on the team if he busts,but then again you are on the hook for him anyway if he is bad and even if he is good you probably prefer getting 3 peak years and then be off the hook over getting 5 really good years, 2 ok ones and 3 terrible ones.
Realistically most surplus value will be created in the first 3-4 years.
Getting 3 prime years of stanton is huge, 3 years is a lot, top10 overall prospects get traded for 3 years of a superstar.
Now as I said the one sidedness of the opt out isn’t ideal, a team option would be the best, but especially the back loaded nature of the contract makes an opt out really not worse than no opt out.
I mean stanton makes like 77M the next 3 years but if he opts out of 200M at age 30 he likely has produced at least 14 WAR those years which is 126 m at 9 m per win.
That is 50M of surplus value which is a top5 overall prospect.
The bigger risk is that he is worse/injured in that timeframe and does NOT opt out.
If he is really good you maybe lose 2 slightly positive surplus value years but you are off the hook for a lot of negative surplus value years.
The question you have to ask is what would you prefer:10 years of him for 300M or 3 years for roughly 80M? Maybe the 300 isn’t that bad if inflation continues and he stays reasonably healthy but 3/80 is pretty good if he is good enough to consider an opt out.
If, at the time of the opt out, the player accurately determines that they are worth more than the remaining value of their contract and so they opt out, this is never a good thing for the team. The team has lost future potential surplus value. Craig’s article is basically saying that the opt out limits that potential surplus value for an acquiring team. You make an excellent point though that if he opts out it is only because he has produced a lot of surplus value and that is great for the team. The chart titled “Stanton Surplus Aging Normally” clearly shows the large surplus that is generated even if he does opt out. So your point: three years of Stanton certainly seems to be hugely valuable is spot on in my opinion. I would also add that a high payroll team that is in a current winning window like the Cubs or Red Sox might not have a lot of other ways to make huge upgrades and might simultaneously not care quite so much about having a less than desirable contract on the books (if it turns out that way) in the mid 2020’s when they are potentially in a down swing anyway (like the current Phillies)
Disagree. First, ‘surplus value’ has different meanings for different teams- see the recent work on the win curve. A player can have negative surplus value for most all teams but positive surplus value for one team- in which case, the trade value of the player is limited given limited suitors, the opt out for the player is rational and also beneficial for the club losing the player.
Second, the orthodox ‘opt outs are never good for the team’ dogma assumes an efficient free agent market. If the free agent market is overvaluing wins- and the current FA rate of $9m/w does appear to be nearly a 50% over valuation- then it can make sense for players to opt out despite having actual negative surplus value. Much of the best work on this site for years has been focused on how inefficient trade and free agency markets are- the dogma that the opt out always hurts teams is anchored on the assumption of efficient trade and free agent markets.
Third, and related, the assumption that ‘if a player has surplus value either in actual win-value or in free-agent win-value, then the club is losing a trade opportunity’ is also dubious because it assumes an efficient trade market and fails to consider uncertainty. The uncertainty related to when a star player will decline is a very tangible uncertainty that plays a key role in most all terrible contracts. The transaction costs associated with trades can often mean that theoretic trade value has actually zero real market trade value. Owners and Execs can and have ruined their careers by trading popular stars on the decline for positive value- this leads to inefficiency. Agent power and player power, as well as PR dynamics, can have a large distortion effect on exactly these types of decisions -when to part with a high performing star seemingly set for painful decline in the not-so-distant future- and opt outs help give teams holding these star player rights an out. It is not rare for teams in pro-sports to ruin entire seasons catering to star power, the cost in mega millions- to ignore this distortion dramatically skews the analysis.
Fourth, agents are not efficient in the advise given to clients, and from markets such as Qualified Offer markets, it is demonstrable that high performing players err on the side of confidence when making contract decisions. Agents and players like to test the free market moreso than is rational. Therefore, even if the opt-out in a vacuum is better for the player than the club, it can become better for the club than the player if, during negotiations, the player/agent over value the benefit of the opt-out.
So long as the star player performs well in the early part of an opt out contract, there exist many reasons to think the opt out will be a net positive for the club despite possible trade opportunities (which as we can see with Stanton, can by hypothetical, inefficient, rife with uncertainty, rife with transaction costs and subject to limited demand due to the high contract value of players with op outs)
I think you’re looking at it like a FA evaluation, not a trade scenario. I think underlying the article is the assumption that the acquiring team has to give up something in a trade for the excess value. So the opt-out clause will cost Miami a good chunk of what they could have asked for in a trade.
Oh yeah, I agree. The reasons to start out by treating it like a FA acquisition though, are: 1) it informs you on what you might be willing to give up and/or how much money you ask Miami to kick in when they ask for more than you want to give up, and 2) Dave Cameron has been saying right along that no team will, or should, take that whole contract even if it was given to them for free; so addressing the question “Would I be willing to take this contract if there was literally no acquisition cost?” is a relevant thing to wonder. And analysis like this suggests that Dave may be wrong.
Since he already has the opt-out, yes, the acquiring team has to be happy if he ends up opting out. That’s because the alternative is that he can’t get a better offer on the free agent market, which means he isn’t worth what he’s opting into.
All other things being equal, the opt-out can only hurt the team. There’s a non-zero chance he plays well enough the next 3 years to be expected to be worth more than $204/7 years or whatever is on the other side of the opt-out, and the team that holds the contract then loses out on that expected surplus value.
Of course, if a team can make a smaller guarantee in exchange for the opt-out then there’s no problem, and that’s generally what’s happened. But we’re talking about a player who already has a contract. In that sense, we have to mention the fact that Stanton has an opt-out, and the fact that it by definition negatively affects the value Stanton’s contract offers a team.
Nearly every relatively poor team (nearly every team actually) is risk averse to 100m+ contracts running into or past the mid 30s. The few teams that aren’t, generally want to lower their expected variance. The idea that – if Stanton opts out- the club holding his rights automatically has lost some trade value or projected surplus value that they would have kept minus the opt out ignores risk aversion, variance preferences, and the inefficiencies of the trade and free agent markets.
Further, the ‘lost trade value’ analysis overlooks ‘no trade clauses,’ which, as with Stanton, can dramatically reduce the liquidity in trade market value. Nearly all optout candidates have sufficient stature to obtain no trade clauses. Additionally, comp picks for lost free agents seem to be left out of the opt out analysis- one of many ways in which the analysis is skewed towards the ‘opt out always good for player bad for team’ dogma.
Well this is exactly what I was going to try to do this afternoon. I guess I can find something else to do now. Well done! I did have the same question though: if we have a large sample that gives us our aging curve why wouldn’t we set “ages normally” at like 70%, “ages poorly” at 25%, and “ages well” at 5%? It seems like that’s pretty important because you come to a vastly different conclusion as to his overall value.
For sure. You can see the 100% ages normally puts the surplus at about $50 million, so any switch from poorly to normally is going to move the number around. It’s not going to get it above $50 million unless you change his 5.3 WAR up or down, and even if you move it way up, you are still going to see your upper limits capped. Even if he is an all-time great the next few years, your upside is only about $120 million, which is great, but the chances of hitting it are pretty low.
Craig, LOVE your work! I always look forward to seeing what you’ve been working on. Two more comments on this piece: first, we can’t discount the possibility of a lost year through injury in ’18 or ’19 along with two 5+ WAR seasons between ‘18-‘20 that also lead to him opting out. That would be unfortunate for the acquiring team. Second, according to your “Stanton Ages Normally” chart, you peg it at a 60% chance of a team deriving at least $73M in surplus value and a 24% chance of deriving at least $110M in surplus value (if he ages normally). Given how hard it is for elite teams to get better that seems like a decent bet to make (especially when there is another 28% chance that the contract is somewhere between meh -$20M and pretty excellent +$53M). One could run the same analysis on Darvish or JD Martinez and it might become quite clear that Stanton is actually quite a good gamble (given that you have money and are trying to get better and are in a win now window). The Cubs, for example, might actually be better off with acquisitions like Cobb and Morrow + Stanton, moving Heyward into the Jon Jay role, than the more orthodox: sign Darvish and Wade Davis and leave the OF as is type moves.
If I were a team I feel just looking at the contract without opt-out is perfectly fine. If for whatever reason Stanton is so good the next 3 years he’s willing to opt out, then he’d be worth a top 10 prospect and maybe more in any trade. Currently, all proposals have not come close to anywhere near that kind of price.
To me the opt out is a bonus. If it happens it means no long term commitment and you already get the prime of Stanton’s career for dirt cheap (money wise and prospect wise). The only thing to worry about IMO for the trade is the likelihood that he opts in. Because that’s the difficult part, needing to project his age 30+ decline phase of his career.
Stanton would likely opt out today if he had the choice, and is it unlikely, through far from clear, that he is worth a top 10 prospect now.
Trade markets are not efficient, particularly with high performing star players approaching their decline phase.
Btw I’m not sure the value with opt out is correct, you probably have to consider that he only opts out when he produced a certain value.
I don’t think he would opt out unless he has produced at least 40m of surplus value because if he produces less that affects the projection and lowers his future surplus value.
Also is 5.3 wins really the 50th percentile outcome? He exceeded that only 3 times in his career and 15-16 he was well below that. Granted with some injuries but with him you probably have to expectconsider him missing 20-25 games a year because that is what he did historically and injury proneness doesn’t tend to get better with age.not sure the projections weigh that risk enough.
This is the first article I’ve read about Stanton that really clarifies why the market thinks Stanton’s contract is roughly market value. A very necessary article.
Stanton opting out is the most desirable outcome. Should he opt out it’s likely he put up 20 WAR over 3 seasons.
The risk is he falls well short of 16 WAR over 3 years and does not opt out, making it likely the remaining years will be at a loss.
Those betting on Stanton putting up 20 WAR over the next 3 years will likely try and trade for him if they are in position to win these 3 years. They will offer the equivalent of 80 million or so surplus value in either cash and/or prospects. Insurance will have to take care of the risk he gets injured and prevented from opting out, or a lower offer to cover the risk
Interesting piece, and well presented. If I can disagree slightly. I don’t think Stanton opts out, even if he’s playing well, unless he doesn’t like the team he’s on at the time of the opt out. He will be 30, he’s promised 30M AAV over the next 7 years, and he’s never been a durable player. But even if he does opt out–I think as an acquiring team, I’d like that–lessens the risk of the contract substantially. But I agree with Craig, the opt-out does lower my offer, especially in terms of what I’m willing to give back in talent.
I really think whoever trades for Stanton is going to regret it. Best case scenario, you get an elite player making $26 M a year for three years in his prime and get a decent amount of surplus value, probably a slight gain from whatever prospects you gave up.
Worst case scenario, however, FAR outweighs the best case scenario. Either a skill regression or injury could erode his talent, making him choose to opt-in to the black hole years of his contract.
Trading for Stanton really only makes sense if the Marlins agree to eat a sizable amount money on the back end of the deal, limiting the financial risk, while ALSO not requiring you to give up major prospects. I don’t see that happening.
If I were the Marlins, I would bet on Giancarlo by agreeing to pay a significant portion of post-option years to mitigate the risk for the acquiring team, and encourage them to want the surplus value pre-option portion.
If he plays well enough to opt out, you make out well. Might even be easier to sell it to him that way, because you’re betting on his talent.
I think that the Marlins’ discussion to sell Stanton on trade includes playing hardball by telling Stanton and his representatives something like:
“We are committed to pursuing a multi-year rebuild during which we will cut payroll and focus on rebuilding our poorly-ranked farm system. Plan A is to trade you to a team that is better positioned to put a contending team around you, particularly for the next three years prior to your opt-out. If you are going to exercise your right to veto a trade – which might also include restricting your acceptable destinations to such an extent that we don’t like the trade package available in return – then our Plan B is to trade Yelich, Ozuna, Gordon, and possibly others, perhaps packaged with some of our underwater contracts such as Chen and Prado. In short, the reality is that refusing to waive your NTC means that you spend the next three years as the centerpiece on a low payroll non-contender here in Miami.”
Maybe Stanton – similar to reports that Votto likes staying in Cincinnati – likes living in Miami enough that he doesn’t mind staying there for the next three years as a star player on a pretty bad team. Or, to be more precise, that Stanton prefers that situation in Miami to accepting a trade to some contenders, such as St. Louis or Boston. The reality that I assume Stanton and his agents realize, however, is that the Marlins aren’t going to bring back a team as good as the 77-85 team that was around the fringe of wild card contention in 2017. The Marlins are cutting payroll, and their minor league system is so poorly-ranked that it’s tough to see meaningful reinforcements coming from that system in the near-term. If Stanton and his agents somehow don’t understand those points, I’m highly confident that the Marlins front office will lay out the reality of the situation.
Exactly this is a reason why “opt-outs are axiomatically negative value for the team” may not be true. If the team can effectively drive the player to opt out by threatening to waste his career…
Btw if the opt out is an issue for an acquiring team the marlins could just offer to eat like 15-18 milllions in thosr first 3 years. That brings his AAV down in those first 3 years but at the same time is not a huge commitment by the marlins.
That means the team gets him for 20m per year those 3 years and he creates more surplus value in the front.
This still leaves the risk of him becoming pujols but that risk exists with or without the opt out.
If the new owners are serious about not becoming Loria 2.0 they should invest those 15M. Three guaranteed years for 20M would make the deal more attractive. That still leaves the risk of him being bad and opting in of course.
We may very well see the real market value of the opt-out provide a large benefit to the Marlins. That is, they may well keep Stanton, get a few more years of excellent production, and then lose him before his decline years because one and only one team (Pujols precedent) signals an intent to wildly overpay. It may also be that, given ownership instability and roster flux, Stanton was only willing to resign with the Marlins with an opt out.
The argument based on ‘lost trade value’ assumes at least 2 teams bidding high for Stanton’s services- but this is only 50% as likely as 1 team signaling a high bid in free agency. The current Stanton scenario is illustrating this dynamic well- the Giants and Cardinals aren’t offering very good prospects, in part because they don’t have them, and in part because there aren’t many bidders for Stanton due to luxury tax reasons. One could obviously argue that this limited trade demand is due to or influenced by the opt out- but it is also easy to see how high priced stars approaching a decline phase may always have limited bidders in the luxury tax age. Given the uncertainty regarding decline and high risk factor, it is not clear that there would be substantially more trade suitors minus the opt out.
If the opt out is a benefit to the player, we should see clubs offering to buy players out of their opt outs- which would indicate the real market value of such clauses. As of yet, I don’t believe there has been such an offer, indicating perhaps a high degree of uncertainty or disagreement on the value of such clauses. Opt outs are simply player options with far greater uncertainty.
No, no, one thousand times no. Players aren’t bought out of their opt-outs simply because MLB contracts are virtually never modified once signed, other than extensions that tack on additional years at the end of the contract.
The far greater evidence is which players we see getting player opt-outs: almost always free agents signing big-money contracts, indicating players with substantial negotiating leverage. The basic principle behind an option is that the option provides value to the holder of the option, who gets to decide whether or not to exercise the option. Player opt-outs provide optionality, and therefore value, to the player. Similarly, club option years provide optionality, and therefore value, to the team. The amount of value depends on the specific terms of the option – in the extreme, a hypothetical club option to keep a player for 1 year at $80 million is such a high price that it simply has basically zero value – but there’s no real question as to which party, before the fact, gains value from the option compared to if the same contract didn’t have any option.
Very poor logic imo.
To say that players aren’t bought out of their opt outs because -normatively- stuff like that just doesn’t happen is poor logic and counterfactual. Players get bought out of certain options – like arbitration options- all the time. They also get bought out of trade clauses- which are an option to reject trades.
The rest of your argument is simply that, by your definition, options are good for players because that is what options are.
But besides avoiding the specifics mentioned in this thread -i.e. that no trade clauses limiting the liquidity of trade value; that agents and high performing players erring on the side of testing the market; that there is a higher frequency of huge, paradigm changing free agent bids compared to huge trade bids (again because of agents); that clubs have preferences for lower variance and risk; etc- it is also incorrect that ‘options’ are always good for players because they ‘provide optionality.’
The value of an option depends on the parameters of the markets. The option to go to arbitration – instead of accepting an offer – depends on the efficiency of the arbitration panel and the criteria that lead to compensation in that market. Players who go to arbitration do not always benefit from that choice- often they make the mistake of betting on themselves when they should not have, in part because of the limited efficiency of the market they opted into.
The opt-out in question – Stanton’s- is one that provides a difficult challenge to him and his agent, because of the many years it covers. Consider that the author of this article writes that, at one point this season, it looked like an opt-out wouldn’t make sense, but then by the end of the season it looks like it would- calculating many years into the future is difficult, full of risk and variance, and the underlying factors that suggest a player should opt out or not can change quickly. A high performing player coming off of a healthy season- with an agent pushing hard to line up a free market mega deal- may very likely over value his ability to maintain health and continue high production.
Clubs giving out options early in long contracts are betting that, because the trade value of a highly paid near 30s star with a no trade clause is limited, the club would be better off if that player accepts a back loaded contract with a moderate AAV in the early years, then performs well in the early years, and then allows their agent push them into an opt out. The entire ‘opt outs are always bad’ dogma depends on the likelihood of the events described above being zero- but recent precedent with the Yankees with both Arob and CC show that the club could have triggered the above event chain by simply letting the players walk into free agency. The fact that some other team would have likely made the mistake (which the Yanks ultimately did) of giving Arob and CC a mega deal- does not mean that the Yankees lost huge trade value, because trade value for hugely paid 30-ish players with no trade clauses is not liquid. Therefore the dogma is too rigid, anti-factual, and wrong- the event chain described above does not have a zero % chance of happening, and so long as it may happen, giving an opt out can benefit a club.
We may well see it happen with the Marlins, as Stanton may not get traded this winter, he may then put up a few more good years for Miami, and then may opt-out CC/Arod style and let some other club (that would not trade much for him because of the non-liquidity of trade markets) make the mistake of buying his decline.
Good analysis.
TLDR: Capped upside, but much larger downside if Stanton turns into a pumpkin.
If he performs well, you lose him after 3 years, after accumulating around 50 MM of surplus value.
If he performs badly you are on the hook for another 200 M. If he turns into a replacement level pumpkin a la Ryan Howard, that is a lot of downside risk to be taking on.
Yes, the capped upside of a potential deal for Stanton significantly lowers the expected value of the deal for the team trading for him.
Also, I bet there is a strong correlation between those who believe that player opt-outs are beneficial for teams and those who believe that front-loading contracts are beneficial for teams.
I have talked to many people that I work with in the industry, other scouts, player personnel execs, and GMs, about Stanton’s trade situation and they all feel the same way. Stanton’s opt-out in no way affects his trade value negatively. Their reasoning is that the opt out is a positive for the team trading for him because if Stanton has 3 seasons that are so productive that he feels he will get more than $218 million, then the team that traded for him will have received surplus value. We could go into the numbers, but its simple really, 3 seasons of 50 home runs and above average defense is worth a whole heck of a lot more than $25 million per season. Their worry is not the opt out itself, its that he may not opt out and they are stuck with Stanton from age 31 to 37 and owe him $218 million. It’s having to pay for his declining value with increasing dollars that worries them.
GMs talk to scouts from other teams about player valuation and say what they really think?
I have a problem with the linear WAR%’s in this article. The idea that there is an equal chance of him producing 8.8+ WAR as there is at him producing a 1.9 or less WAR is kind of questionable isn’t it? Not a big deal though.
My bigger question goes back to the starting point with Sarris’s article saying that Stanton is worth 400 million over 10 years. My biggest complaint is the idea that Stanton at 30.5-30.11 will produce the same WAR as he will at 28.5-28.11 season. Players are aging worse lately. If you lower his first 0.5 drop by just 1 year he will produce 4 less WAR overall and over 40 million over the life of the contract. That is a sizable difference and I believe a much more accurate expectation in this day and age. I would be interested in seeing these numbers, even though I think 10-360 would still be too much money for Stanton in FA.
Can the Marlins offer to pay down the contract in years 4-10 IF he opts in? That appears to be where the risk is and what is actually driving down his price. That would leave the acquiring team with two scenarios:
1. He opts out. You most likely received significant surplus value. Bummer that he leaves (or you have to renegotiate) but you paid well under market prices for 3 years.
2. He opts in. You may or may not receive surplus value but your risk is mitigated by the Marlins paying down 7 years of the deal at say $5MM/year.
A deal like this would guarantee that the Marlins land a package of talented youngsters since the risk is minimized to the acquiring team. Even if things go bad, being on the hook for $5MM annually in dead money from 2021 to 2027 isn’t the end of the world for the Marlins.
Sorry if I missed your comment on this, but you can only use linear values for teams if they are enormous financial entities (like reinsurance firms worth $10s of billions). Otherwise they should be risk sensitive and, all else being equal, preffer a shorter contract with the same surplus but less risk.
There is also the risk that TV and radio revenue go away and they’re holding the bag on these huge worthless contracts.
Dave’s argument that opt-outs are always bad is simply financially ignorant, by the exact same logic arbitrage would always be bad for any investor.
Lowering long-term expected profit but gaining lower variance, greater flexibility and lower risk, can be a substantial benefit benefit.
I am not seeing the arbitrage opportunity for the team here. It’s more like someone holding an option to buy your $30 stock for $50. The option is not going to be exercised unless your stock goes over $50. In that case, without the option, you could sell the stock yourself, so you lose the surplus value of the stock. There is no scenario where that option benefits you.
Where is the lower risk? If the player is good, you lose his surplus value. If the player is bad, you are stuck paying him. Opt-outs are the classic “heads I win, tails you lose” scenario. Dave is right.
This is not true.
Players/agents make a mistake on their valuation all the time- it happens frequently in decisions of whether or not to go to arbitration.
Further, trade markets for highly compensated players with no-trade clauses are not liquid. Therefore, in your analogy, you cannot simply ‘sell yourself,’ – the opportunity to cash in on surplus value via trade is extremely restricted.
Recent precedent with Arob and Sabathia shows, definitively, that there is a scenario where the option could benefit the club. That is the problem with this ‘zero percent chance a club could benefit’ dogma- it is too extreme. The Yankees were in the clear with a great situation with both Arob and Sabathia- neither could be traded for much in trade value, both of their contracts were actually underwater or near underwater with huge risk factors, and the Yankees had a great chance to lock in the value that had already accrued by divesting themselves of these risky assets. That they chose not to is simply another example of how common misvaluations are with such long-term calculations, agent power, and star power involved- these misvaluations are what can make opt outs a benefit for a team much more frequently than the zero percent you are suggesting. With players like Sabathia, even a decent trade offer would not be as beneficial of being rid of the contract and decline phase.
All the ‘opt outs are never good for clubs’ arguments depend on almost perfectly efficient markets, where player valuations never happen. This site came to prominence in large part because of the great commentary it offered on how inefficient free agent and trade markets were in MLB- and how frequently clubs ‘bought the decline,’ and how these inefficiencies could be taken advantage of by smart GMs. The opt out is simply another wrinkle to exploit exactly this dynamic.
So the opt out advantage scenario is that the player may make a mistake in calculating his future value? That’s a pretty thin thread to hang your assertion on.
Of course, people make mistakes. Evaluating the future value of a player (correctly) is hard. But there are going to be many people evaluating the future value of the player at that time, so he is likely to have a pretty accurate idea of what his market value is. If it is higher than the value of the remainder of his contract, then he leaves, and the team loses. If it is less, then he plays out his contract, and the team loses (probably. That evaluation might be wrong too!).
It’s very annoying to read someone throwing around insults like “financially ignorant” while simultaneously misusing terms and concepts.
An opt-out does not provide “greater flexibility” for the team. The opt-out is solely at the discretion of the player, so it provides him, but not the team, with greater flexibility. The team has still guaranteed Stanton the money after the opt out if he chooses to take it. Speaking in terms of multi-year budgeting and planning for a team, it can also meaningfully reduce flexibility because, for example, after 2019 it may appear very unclear whether or not Stanton will opt-out after 2020, which impacts a team’s thoughts on additional multi-year free agent signings – will Stanton’s money be on the 2021 books or not? – as well as decisions such as thinking about whether it will have an opening in RF for a minor league outfield prospect after 2020.
It also doesn’t lower the team’s risk, because risk is defined by downside. The team still has the full downside of post opt-out years if a player underperforms or has a serious injury prior to the opt-out.
To the extent that there’s any “lower variance”, it’s the sort of variance that a team shouldn’t want because the scenarios that are precluded by the opt-out are largely positive scenarios for the team, where Stanton ages relatively well and thus outperforms his contract after the opt-out.
I did not misuse any financial term.
The opt out can provide greater flexibility, as a player with an opt out has a lower projected cost and years commitment. This means that for any player on a 10 year contract at 200m, with a 4th year opt out, the club can be certain of getting the first 4 years at a given price, but is projected to get have less commitment with respect to the last 6 years and the remaining price.
Your argument depends entirely on the idea that the opt out decisions will always be efficient- and also- that the trade value of a near 30s, highly paid player with a no trade clause is liquid. If either of these things are not true, your argument falls apart. If players/agents frequently make inefficient decisions and opt out when they shouldn’t, then teams can clearly benefit. If near 30s, highly paid players with no-trade clauses have theoretic trade value (i.e. surplus value) that, by reason of limited bidders and the no-trade clause, does not materialize into actual trade value, then your argument falls apart.
First, you misused the concepts of variance and risk. I explained the latter in my prior comment. As for the former, claiming that a lower variance investment is necessarily better than a higher variance one requires the assumption that the expected value of each investment is equal. As Craig lays out in the very point of this post, that’s not the case if we like at two contracts that have identical terms other that one including a player opt-out, because the opt-out precludes some outcomes that have a highly positive return for the team.
As for your focus on no-trade clauses, it need not be the case that a team realize surplus value via a trade. The team may simply wish to keep the player for the remainder of his contract, but not be able to do so because of the player opt-out.
Additionally, you seem to ignore the many cases when a player does in fact agree to waive a no-trade clause. Taking just a couple recent examples, Verlander waived his no-trade clause to go to the Astros, and McCann also did so. Alex Rodriguez did so when he was traded from the Rangers to the Yankees, as an example where a player was traded with a contract with a very large amount of money remaining. Given the dynamics of an opt-out situation – the player is a free agent deciding where to play, and the team that signs him is offering an attractive enough combination of money and situation that he agrees to sign there – those are situations where we can reasonably be most likely to expect that a satisfactory trade could be found and that the player would agree to the trade.
Another reason the ‘opt outs always bad’ dogma is off the mark is that is assumes players will only opt out if the market will offer more AAV, whereas (though not likely in Stanton’s case) a player may opt out for similar or even slightly lesser AAV to get more years.
Factors that can effect the value of an opt out include: when the opt out occurs, the age of the player involved, how many opt outs, etc., An early opt out right before projected decline, where the player has to make a difficult decision on a ‘many-years’ player option, where the player (and critically the agent) may prefer to go to market in order to add years (even if not able to add AAV), can be regularly beneficial for teams: especially if the player has a no trade clause or salary that would restrict the liquidity of his trade value.
I think what you’re missing here is the probability of catastrophic career-ending injury. I don’t know what that looks like, 5%? 1%? But whatever it is, you’re missing it, and the fact that it could result in 0 WAR for the whole contract (even with insurance – which isn’t priced into this), has a huge effect on expected values.
If I was an NL team I would be concerned that Stanton might not be a viable OF the last few years of his contract. And even if he is, I wouldn’t want a 6’6 250lb+ 35,36,37 yr old, banging about in the OF.
I’d think NL teams would prefer Stanton stays healthy and productive and opt’s out rather than mess with those last 3-4 years.
Late to the party here, but you simply cannot project past 2020 using 2017 numbers. As of now, if Stanton opts in, his projected WAR for 2021 is based of of his contract extension. Go backward from $218/7yrs with all of the depreciation and find what his maximum value “resets” to after the next 3 years.
I plugged that into the Stanton Salary/WAR worksheet and the break-even point indicated a 4.4 WAR in 2021. So, one theory would suggest Stanton would opt-out if he’s projected for worth more than 4.4 WAR in 2021. However, how many 31 year old, 4.4 WAR players get $218/7yr?
Bobby Cano had 5.8 WAR in his age 30, pre FA season and got $240/10yr – and that was ridiculed as an immediate albatross.
Chris Davis had 5.7 WAR in his age 29, pre FA season and got $161M/7yr – also seen as a dunder-headed deal for the team.
Cespedes had 6.7 WAR in his age 29, pre-FA season and got $75M/3yr (with opt-out).
Basically, recent history shows that a 31 year old player coming off some great numbers will NOT likely be paid $218/7yrs. A player with a mixed history like Stanton’s who projects to 4.4 WAR at age 31 does not likely qualify for a long term megadeal at all. If Stanton does not opt out, the team is absolutely upside down by A LOT. It would likely also indicate that the team did not receive all that much surplus in 2018-2020.