December 15, 2014
The Surprising Math Teams Use to Value a Compensation Pick
All other factors equal, it is preferable to sign a free agent who is not attached to a qualifying offer than a free agent that is QO-attached. This is obvious. What is not obvious is by how much, or whether that “how much” is always constant from free agent to free agent for each team. First round pick protection, competitive window, payroll limit, state of one’s minor-league system, and upcoming draft class will all determine how each team costs (values) losing a draft pick. According to traditional financial, economic, whatever-you-want-to-call-it theory, weighing these factors all makes sense. But, according to traditional theory, each team would individually weigh the cost of giving up a draft pick equally across all QO-attached free agents. Example time:
Team X puts the cost of losing its first round pick at $8M. Absent the qualifying offer, Team X values Max Scherzer at six years, $150M and Ervin Santana at two years, $25M. Taking the cost of losing a first round pick into account, Team X should value Scherzer at 6 years, $142M and Santana at 2 years, $17M.
Cool, makes sense. Alas, mental accounting, which posits that “people spontaneously generate their own mental accounts, and where we place these boundaries subtly (but profoundly) influences financial decision making,” indicates that our traditional theories may be oversimplifying things here. Specifically, it notes that we create topical accounts, in that our decisions are altered by the context of the situation. Whereas most think it absurd to drive 15 minutes down the road to another car dealership to save $75 on a $25,000 car, many will stand in line for an hour in the middle of the night to save that same $75 on a $250 smart phone.
Question: What does this have to do with QO-attached free agents?
Answer: Given the use of topical accounts, we could hypothesize that if GMs categorize Scherzer as an impact player and Santana as a role player—and are less willing to give up a draft pick to get a mere role player—that teams might be either (i) undervaluing the cost of the QO when valuing top free agents (saving $75 on the car) and (ii) overvaluing the cost of the QO when valuing lower-end free agents (saving $75 on the smart phone).
We have a hypothesis, but we did not want to stop there. We set out on a journey to test this hypothesis. A consultant might pitch the hypothesis testing that followed as a two-pronged attack. The first prong was looking for mental accounting within the words of the decision makers (asking general managers how they value QO-attached free agents) and the second prong was looking for mental accounting in the actual decisions (contracts) of the decision makers. Ultimately, we want to see if what the decision makers are saying is at the very least indicated by their actions.
Many decisions makers were emailed and a few very generously (given my role in the universe and the winter meetings) responded. We did, however, hear answers along the lines of, “Yes, it is easier to part with a pick when the return is a star player than it is when the return is more of a role player.” Unfortunately, there are not enough data points (only 22 qualifying offers rejected in the 2013 and 2014 offseasons) for proper analysis. Moreover, the player mix of the dataset presents problems. Some examples: The best QO-attached pitcher signed to date is probably 2013 Hiroki Kuroda and only two QO–attached middle infielders signed, Robinson Cano and Stephen Drew. Lastly, it is tough to evaluate this offseason’s signings with some players still unsigned. All this means that an outlier signing or projection can overly sway the analysis, while making such an outlier potentially impossible to identify. As a result, testing the previous hypothesis with the current info is probably both foolhardy and reckless.
So, that is a dead end (for now!), but all is not lost. Our quest was provided a northern star from a general manager who mentioned that he is not sure if teams discount the loss of a draft pick for top free agents because of the previously described mental accounting, but that teams “probably do not mind giving up a later first round pick if they were getting one back in the supplemental round.” The neat part of this observation is that it is mental accounting manifesting itself in a different way.
New hypothesis: Teams are more likely to sign a QO-attached free agent (and sacrifice a first round pick) if they are going to get a different pick back in the compensation round the same year.
Question: What does this have to do with mental accounting? In other words, is this not just common sense?
This hypothesis seems to be backed up by the numbers (the language is cautious because of the sample size). Of the 22 QO-attached free agents, 10 were signed by teams sacrificing first round picks—but only five of those 10 resulted in a team forfeiting its only first/supplemental round pick. See for oneself below:
(A note on the five teams that forfeited a first round pick to sign a QO-attached free agent: All but the Nationals had multiple first round picks* in the prior year’s draft.)
On the flipside of this mental accounting coin, of the 13 non-QO-attached players who signed contracts worth more than $30M (for instance, star players who were traded midseason and were ineligible for a qualifying offer), six were signed by teams that would have forfeited their only first or supplemental round pick—if the players had been QO-attached.
· Of 22 QO-attached free agents, only five were signed by teams sacrificing their only first round pick.
· Of 13 non-QO-attached free agents, six were signed by the same class of team.
While it is not enough to confirm, all this does indicate that teams place some amount of value on avoiding an early-round-pick deficit. The new draft caps and slot limits probably add to the emphasis placed on early round pick balance—that is, on not having zero picks—because no longer are teams just losing a pick, they are decreasing their bonus-slot leverage for the entire draft. Furthermore, we also see indications that teams are somewhat lumping the second round pick into their trade balance mental accounting. Why? Because only once have we seen a team forfeit all of its picks in the first and second rounds; that being the Orioles when Nelson Cruz’s price dropped to one year, $8M deal with incentives.
Status Check: Through the indication of front offices and their actions, we have reason to believe that there is some sort of premium placed on not forfeiting all of one’s top draft picks. Let us see what is happening so far this season to see if it is consistent with that thought.
Twelve players received the QO this winter, with only two (Max Scherzer and James Shields) yet to sign. The 10 QO-attached players that did sign, did so as seen below:
So far, 2015 looks a lot like what we saw in 2013 and 2014. Two (Mariners and Mets) of 10 teams signed players that cost them their only first round pick; one, the Mariners, had multiple first round picks the year prior. Given all of this, it would be surprising to see Scherzer or Shields sign with a team sacrificing its last remaining first round pick; even more surprising if such a team did so without having multiple top picks in last year’s draft. This is a bit of hedge, given that many of the top market teams either have both a first and a supplemental round pick (LAD, NYY, DET, SF) or are protected (PHI, BOS, TEX). Three potential movers that would be forfeiting their top pick would be the Angels, Nationals, and Blue Jays (who had two picks in the 2014 first round).
Additionally, the impact of mental accounting might potentially go beyond indicating which teams are likely to be in the market for certain free agents, and this goes back to the previously mentioned premium being placed on conserving top picks. It has been written before, but a premium or a preference or any sort of overarching theory that affects valuation beyond the actual value provided by an asset (player, pick, cash, etc.), can cause suboptimal decision-making. Or, it can cause us as analysts to be forced to ask why those in the game believe such decision-making to be optimal.
In any case, the fact that giving up picks seems to reduce demand by taking teams out of contention for QO-attached free agents means that there very well might be a discount on some of these free agents. Back to the front office voice: “The qualifying offer attachment can reduce the market, lowering the cost of the free agent.” In other words, the mental accounting being exhibited, if it does exist as it appears to, can reduce the chances of winner’s curse for teams. Winner’s curse states the value of an asset will most likely be the average of the valuations of all parties involved in the bid and, therefore, the winner is likely to overpay. The corollary to this is that the fewer parties involved in the bidding, the less pronounced the effects of the winner’s curse. Therefore, the fewer teams involved in the bidding due to the trade balance mental accounting, the less likely an overpay. It is important to note that this does not carry across all QO-attached free agents because if enough teams have extra picks to sacrifice or picks protected, then demand will not be reduced enough to drive down price. Demand being driven down seems to happen more with the lower-tier free agents (Lohse in 2013, Cruz in 2014, Cabrera in 2015), where there is not a huge market for their services. Again, these are further questions and things to watch for rather than conclusions, but once we have enough of a sample size to take a crack at values and overpays, we will have a better idea where to start.
*First round picks, except where specified, refer to first round, compensation and supplemental rounds.