The New Economics of the MLB Trade Market
| 38
|
by user DNL
Michael Lewis' Moneyball was the capstone on the sabermetric movement. Many think that the thesis of the book can be summed as this:
- Ignore "batting average" and "RBI." Downgrade stolen bases. Focus on "on base percentage" and "slugging."
But, really, that's not what the book is about. Moneyball's central premise is that many times, the marketplace fails to put the correct value on players, and the effective GM must identify these inefficiencies and take advantage of them.
One of these inefficiencies was caused by free agent compensation. Major League Baseball teams that lose players to other teams via free agency may be entitled to compensatory picks. The system is on the complicated side, but I'll boil it down like this:
- Players are given a score by the Elias Sports Bureau between 0 and 100. Anyone in the top 30% at his position using this scoring system is considered an "A" player.
- If a team from the previous season signs an "A" player, that team forfeits its first-round pick in the next year's draft to the player's old team, if the old team offered the player arbitration. If a bottom-15 team from the previous season signs an "A" player, that team forfeits its second-round pick in the next year's draft to the player's old team, if the old team offered the player arbitration.
- If a team loses an "A" player, that team also gets a "sandwich" pick in a round that is (as the name suggests) sandwiched between the first two rounds.
The above isn't exactly right, but it's close enough to paint a picture.
Class "A" goes deeper than one would think, and when you add on Class "B" (which gives a first/second round pick, but no compensatory pick), you're looking at 50% of the Majors. Check out Baseball America's Q&A discussing how last year's compensatory scheme worked out, and you'll see that a lot of picks changed hands.
Billy Beane, as suggested by Lewis, was one of the first to truly understand and abuse the system. Take, for example, his July 25, 2002 deal to acquire Ray Durham from the Chicago White Sox, for ne'er-do-well pitcher Jon Adkins. Beane took on some salary -- Durham was due over $2 million for the remainder of his contract, but the White Sox sent cash with the 2B/OFer -- and gave up a token. On its face, a fair deal. But in reality, it was a steal, as Beane also got two draft picks when Durham -- an "A" free agent -- signed with the San Francisco Giants.
Look at it this way:
- The A's gave up a schlub and some money
- The White Sox gave up Durham.
- The A's got Durham for two months and two prospects.
- The White Sox cleared salary space and opened up a spot for a guy like D'Angelo Jiminez.
Before you ask why the White Sox would be so dumb, remember that at the time, many thought that the new CBA would eradicate compensatory picks. That being the case, the White Sox were probably acting rationally, and, well, got screwed.
But suffice it to say that many teams undervalued these picks, and had for a while. The compensatory system is basically a price floor when it comes to trading -- or, at least, should be. If not, there's an inefficiency ripe for abuse. Is it being abused? If not, is something else? Let's see.
Do Teams See the Price Floor?
Arguably, from the period that Moneyball examines, the answer to this question was "usually not." Teams would sell their B-level free agents making B-level salaries for pennies, hoping to get something for the player before the player left of his own volition. Rob Neyer once argued that Beane was uniquely exploiting this by buying up players at this bargain, and then getting compensatory picks in return.
Neyer's article, from what I can gather, was written almost four years ago. Have GMs wisened up?
Maybe.
It's hard to say, because teams aren't offering players arbitration as readily as we'd expect. Last year, the New York Mets declined to offer arbitration to both Mike Piazza and Braden Looper, and for good reason -- both could have likely gotten more from the arbitrator than from the market. And similarly, there have not been a lot of big name free agents (defined loosely) who have been moved at the deadline. The only team that was awarded a compensatory first-round pick in the 2005 draft for a player they acquired during the 2004 was the Boston Red Sox, from the Los Angeles Angels. The Red Sox acquired Orlando Cabrera at the deadline that year; Cabrera signed with the Angels five months later. And I don't see any players that fall into this category for the year 2006.
The Cabrera deal, then, is the strongest evidence of a lack of understanding, I'd suppose. But there are mitigating factors. First, Cabrera was traded by the Expos, who were owned by MLB -- it's hard to say who was calling the shots. Second, the Expos got two prospects in the deal and one of the Alex Gonzalezes, who they soon after flipped to the Padres. So it's not like they were taken for a ride. Finally, because of the Expos' questionable status, it's hard to say that they could have afforded to offer Cabrera salary arbitration anyway; that is, they may have gotten nothing in the deal.
The only evidence, therefore, of a collective "wisening up" by GMs, is anecdotal. Given that the GM in the Cabrera deal was Omar Minaya, my anecdotal evidence is even stronger: Paul Lo Duca.
This past off-season, the Mets (that is, Minaya) acquired Lo Duca for low-minors prospect Gaby Hernandez. Hernandez was a well-regarded prospect, but, like any other teenage pitching prospect, all promise and no guarantee. If that sounds familar, it's pretty much what you get in the second or third round of the MLB draft.
The reason that matters? The Mets' other options at catcher were Benjie Molina or Ramon Hernandez, both free agents, and both likely to be offered arbitration (which turned out to not be the case, but Minaya didn't know that). As a friend pointed out soon after the Lo Duca trade, no matter which of the three the Mets acquired, the cost in talent would be the same. That is, Minaya was giving up a prospect whether it be by trade (for Lo Duca) or by forfeiting a draft pick (in signing Molina/Hernandez). Am I giving him too much credit? Perhaps. But perhaps not.
But the better anecdote? Post-deadline waiver trading. If the league as a whole, has realized that free-agents-to-be have value, then these players are simply untradable after the July 31st deadline. Yes, they can be traded up until August 31st and still be playoff eligible for their new team, but in that situation, they'd have to clear waivers.
But here's the thing: That can't happen.
Because these players are, effectively, coupons redeemable for two free prospects (all you have to do is pay their prorated salary), there's little reason for the Royals, Marlins, D'Rays, Nationals, etc. to all pass on a waived player. Put in your claim and worst case scenario, you pay $2-3 million for two high picks in the upcoming draft. Honestly, that sounds like a bargain.
Indeed, while individually, teams may have reasons for passing, collectively, one team has to be active on this front for any given player. The compensatory scheme MLB employs should cause an outright bar on these types of players being traded after the 7/31 deadline.
Whether or not the cause, that seems to be the result. Looking through the August 2004 Transactions Log, it appears that no free-agents-to-be who would garner compensation were traded, as expected. Same goes for the 2005 log. Maybe I'm missing something, but if not, at least the theory has some chance at validity.
The Effect of a Recognized Floor
For sake of discussion, I'm going to assume that by now, GMs far and wide understand that having a guy who is going to walk at the end of the year is something of value. (Apparently, otherwise insane fans get it, so one would hope GMs do.) In short: you'll get one if not two picks when he walks.
From the seller's vantage point, they need to be compensated for the two picks we'd otherwise get, assuming these teams could/would offer the player arbitration.
By and large, this means that the seller needs to get talent equal to what they'd reasonably expect to get in the draft. But this is not a simple matter. For small market teams -- I'm envisioning the Matt Bush-over-Stephen Drew-and-Jered Weaver Padres -- the extra picks are worth relatively less than they would be to other teams. In the same vein, the salary relief these Padres would get in dumping a vet would be very valuable, as an extra $3 million in savings could be used to sign Weaver over Bush.
But in most cases, the floor would be more signinficant, and it would require that the seller gets more than just salary relief. And even if it is a case like the Padres, well, there's still the buyer's point of view.
Buyers know that no matter how the season plays out, once they acquire a guy in a walk year, they're likely going to get something (in the form of compensatory picks) in return. Put two buyers in the picture, and suddenly, this benefit gets passed to the seller, as the buyers are trapped into what is akin to an auction.
This being the case, the price for would-be free agents should be higher than it was before the Beane revelation. It's hard to tell if that's true yet, but suffice it to say that Durham-for-junk trades are going to be fewer and further between.
The New Moneyball: New Market, New Exploits
All that said, let's return to the beginning: The marketplace has changed, but there still are some inefficiencies. Where are they? Are they being abused? Am I full of shinola?
Reader warning: The following facts did not happen as they seem. But they make for a good illustration. I realize that Jim Duquette wasn't thinking this bright. Thanks.
On July 30, 2004, Duquette traded Ty Wigginton and prospects Justin Huber and Matt Peterson to the Pittsburgh Pirates (kind of -- Huber actually went to Kansas City for Jose Bautista; Bautista was flipped to Pittsburgh). The Mets got back Jeff Keppinger and Kris Benson. Benson was a free-agent-to-be. The Mets were contenders only in the hearts and minds of leprechauns and unicorns. Many -- people, not fictional characters -- believed that Benson was going to sign with the Mets anyway after the year. The trade received nothing but downward pointed thumbs.
Ignoring the other, totally horrific deal that day, let's boil this one down to two prospects for two months use of a decent starting pitcher -- when the two months don't matter. Terrible? Yeah. However, let's look at it another way: Two months to see if Benson can handle New York, to negotiate a deal with him, etc. In return, the Mets gave up a few million dollars (Benson's prorated salary) and a downgrade in prospects.
What's that? "A downgrade in prospects"? Well, yes.
The Mets, effectively, gave up Huber and Peterson for right to keep their #1 (or, as it turned out, #2) draft pick. That pick, Matt Durkin, isn't doing too well, but then again, neither is Peterson. So, we can truly boil down the trade as simply Huber for a two-month rental/tryout/negotiation or whatever with Benson.
Again, this isn't a good deal. But it's much closer than the usual stance on it. And it gets slightly better. What if the Mets decided Benson wasn't going to work out? Well, they have a built-in insurance policy -- they can offer him arbitration and let him walk. Then, they get a compensatory pick and a sandwich pick. The compensatory pick is, for all intents and purpose, Durkin (otherwise we'd be double-counting). So, they gain a sandwich pick.
Assuming, again, that Benson was willing to sign with the Mets after the year anyway, the deal works out in one of four ways:
- The Mets trade for Benson and decide to keep him. In effect, they give up $2m and Justin Huber, plus whatever the contract for Benson is.
- The Mets trade for Benson and decide to let him walk. They give up Huber and $2m, and get a pick between the first and second round.
- The Mets don't trade for Benson, and decide to sign him. They give up Durkin, effectively, plus whatever the contract for Benson is. They also give up the ability to see if they can work with Benson, and may find that the contract with him is higher than it would be had they not let him hit the open market as hard.
- The Mets neither trade for Benson nor sign him. They give up nothing.
Taking the above, the four choices appear rankable in order of preference, but the gap between the worst (the actual trade) and the best (just signing him, or, more appropriately, not signing him at all) is certainly smaller than many would originally think.
Is this to say that Duquette's Folly (Part I) was not that bad? Hardly. I'm no Duq apologist! The point of this is different:
Sometimes, it's worth making a pre-emptive strike, and grabbing the guy before he hits the free agent market. You get to test drive his services and it doesn't cost you a high draft pick if you decide to make that long-term purchase. Plus, you get to choose which assets you want to spend.
Of course, the marketplace won't allow a pre-emptive strike to occur efffectively at the trade deadline. The market is too hot for guys like Benson (sickeningly), so the premium one pays to get him is too high to make the deal worth making unless he'll give you value in the immediate term.
Therefore, it has to be made earlier.
As in, December.
Potential Exploit: The Three Markets for Alfonso Soriano
Offseason, 2005
In December of 2005, the Texas Rangers ripped off the Washington Nationals, sending the vastly overrated Alfonso Soriano to D.C. in exchange for Brad Wilkerson, Terrmel Sledge, and a minor leaguer.
In and of itself, this deal was horrific. Soriano's OBP was a Rey Ordonezian .309. He crushed 36 homers -- in a ballpark that is batter-friendly. He was moving to a stadium that wasn't so nice to hitters, to a position that mitigated his offensive prowess, and to a team that was not winning any time soon.
And yet, it managed to work out well for the Nats, with Soriano putting up superb numbers for a DH. And that's saying something. Top 10 in the NL in OPS, SB, HR, SLG... yikes.
But that's the value of hindsight. In reality, the Nats traded for a guy at the time not worth anywhere near the $10 million he was going to get. They didn't need a second baseman. They were not winning, with or without him. Oh, and before you say that they were acquiring him for the future, think again -- Soriano is a free agent after the season.
Yet in retrospect, maybe the deal did make sense, even assuming that Nationals GM Jim Bowden really did believe that Soriano was overrated.
Let's discuss what Bowden gave up. The minor leaguer -- Armando Galarraga -- was a 23 year old who split the season between A and AA ball in 2005. He had an ERA of 5.19 and a WHIP of 1.32 in 13 starts. Eh.
Wilkerson and Sledge are both 28. Neither are free agents any time soon, which is a plus, but also a minus. Neither really had a role on this team. The Nationals now have owners and are now a true large-market team. Do you start Terrmel Sledge? Nah. Wilkerson, maybe/probably, but you can do better. But Wilkerson's not going to be part of the picture if and when the Nats are ever good. And if he is, Soriano would be, too.
Oh, he also gave up $7 million -- the difference between Soriano's salary and Wilkerson's.
It's a lot, yes. But how bad it is depends on what you intend to do with Soriano down the line -- or, more accurately, if you consider that in some way, Soriano's fate will be resolved by the end of the next off-season, and each way has a different payoff.
But the fact that the price for Soriano was so cheap -- yes, cheap -- suggests that most GMs don't think that far ahead.
It's cheap because it's mostly cash. Wilkerson is worth something, but again, he's 28. You're not going to see much improvement (if any), so you know what you're getting. Sledge and that minor leaguer (Andres Benitez or something), feh.
It's cheap because Soriano is worth something even if he sucks. That's because there are two more markets for Soriano.
July, 2006
The trade deadline. Bowden has a severe disincentive to trade Soriano. He only costs about $3.5 million for the rest of the season, and if you let him play it out and then watch him walk (with offer of arbitration in hand), you get two draft picks. And thanks to Beane, Lewis, and the popularity of Moneyball, everyone knows this.
Therefore, any demand for Soriano must take into regard the price floor. That's why offers of Brandon McCarthy -- a pitching prospect better than Wilkerson plus Sledge could ever fetch -- make for credible rumors. And a lot of that would be true even if Soriano wasn't on a fifteen week tear. Here's a guy who was traded for A-Rod, virtually straight up; that ESPN called "a rare combination of power and speed . . . one of the most productive infielders in the majors the past four years" when he was dealt to Washington; etc. If he slumps? Bad environment. Doesn't want to play LF. Having a hard time adjusting to the NL. Who knows, who cares -- someone will buy the line. So, you'll not get Brandon McCarthy. You'll get someone not quite as good. But the risk is low, and again, you're burning dollars, not players.
If you don't move him by the deadline, well, market three comes up soon.
December, 2006
If Bowden's interest in Soriano was genuine, and that he truly wanted him as a left fielder, and for the next five years or so. Perfectly reasonable. In that case, he gets exactly that -- and an insurance policy. If Soriano doesn't want to play left, or simply can't, Bowden offers him arbitration and lets him walk. In return, he gets two draft picks.
If he does, maybe you keep him. Perhaps a discount is in order; after all, you have had a full year to negotiate. Maybe not, but in any event, you keep your second-round draft pick. That pick is clearly worth more than Terrmel Sledge. So in effect, you get Soriano and a pick for Wilkerson, Sledge, and a minor leaguer. Oh, and $7 million. (That's the difference between Soriano's salary and Wilkerson's.) Hardly a small amount, but maybe worth the risk.
The real gold, though, is how Bowden managed to turn the money into options. If he thinks the Nationals are close, he already has the big bat he wants, and in the position he wants (LF). If not, he has a trade chip to bring in younger talent -- younger than the 28 years Wilkerson brought. At worst, he gets two picks near the top of the 2007 draft. The price is high -- too high for Billy Beane and others -- but I'd imagine that this could be done with cheaper, similar players.
Imagine this totally ficticious conversation between Bowden and then-future co-owner Stan Kasten:
| Jim: Hey, Stan? Great news... just worked out a deal for Alfonso Soriano. We give up Wilkerson, Sledge, and some minor league arm -- Armando Galarraga.
Stan: ... Jim: Stan? Stan: Yeah, I'm here. I don't get it. We have a second baseman, right? Or are you trading Jose, too? Jim: No, he stays. We'll move Sori to left. Stan: He's never played there before, right? You want to trade an outfielder for an infielder and move the infielder to the outfield? Jim: Well, yeah. He's a much better hitter. Steals bases. Exciting. Stan: And is 30, will cost what, $10 million, and is a free agent after the year and will leave so he can go back to second base, where he's truly valuable. This sounds like a "win now" move, Jim, and we're not a "win now" team. Frankily, Jim, I don't get it. Jim: It's not a win now move. In fact, it's not even a move for Soriano necessarily. Think of it this way. We just moved from Montreal to D.C., which means we have a huge fan base and a big stream of revenue. We can't buy our way into immediate contention, though. So, we need to do the next best thing -- buy for the future. But how do you buy prospects? You can't. Can't trade for draft picks. The best we could do is become active in the market for foreign players, and we'll work on that too, but there's a neat other thing we can do. Stan: Okay... Jim: Right now, everyone sees Soriano the same way you did -- $10 million plus talent for a win-now player. And no one wants to pay that high a price. The Nats want Lastings Milledge from the Mets, but the Mets aren't going to sell the future until they're more confident that they're truly contenders this year. But that will change in July. In July, there are going to be a lot of teams looking for a big bat who can play outfield or second base, who can steal bases, and who won't saddle them long-term with an inflated contract. So, if we're not competitive, we flip Soriano in July for prospects. And note that both Wilkerson and Sledge are 28, so it's not like they are in our long-term plans. Stan: Hmm... interesting. But $10 million is a lot for a prospect or two. Jim: Three point five million, actually. We'd only pay two-thirds of Soriano's salary, because his new team would be responsible for the other third. That's $6.5m. Subtract out the three mil we already owe Wilkerson, and it's $3.5 million. Heck, we'd probably have to pay that to Wilky in 2007. So, it's basically free. Stan: Risky, though. What if you can't trade Soriano? Jim: That'd be bad, yeah, but we just let him walk as a free agent. Stan: JUST LET HIM WALK!?! Jim: Well, yeah. He's almost certain to be a Class "A" free agent, which means we'd get two extra draft picks, each probably in the top 50 of the 2007 draft. Stan: Oh... and that'd cost $7 million -- Soriano's salary minus Wilkerson's. Plus, if we could replace Wilkerson with a guy making the minimum in 2007, we'd save another $3.5 million. So, worst case scenario, we get two top-50 picks for $3.5 million. Jim: Yep. Stan: Make it so! |
It probably didn't happen that way. But it could have. No, it should have. For some reason, the marketplace allows such an absurdity.
Exploit #2: Is Billy Beane Really Shopping Barry Zito?
Perish the thought! At the time of this writing, Oakland is tied for the AL West lead. Plus, Zito will fetch them two draft picks. Would they really trade Barry Zito?
Maybe. Probably. Actually -- yes.
First, are they buyers or sellers? If they're buyers, then there's little reason to think that Zito is going anywhere.
But, they're probably sellers.
| flipping through history | |
|
Beane's typical advantage in July is ever-shrinking. It was derived from a market landscape where teams were desperate to shed money while simultaneously failing to consider the benefit gained when your free agent leaves for other pastures (that is, compensatory picks). Neither of these things are as true, if true at all, as they were even four years ago. Take, for example, the recent trade between the Milwaukee Brewers and Texas Rangers, sending Carlos Lee to Texas. The price was simply too high for Beane to consider (three Major League-ready players) assuming Lee walked at the end of the year; the two draft picks just do not soften the blow. (For Texas, it's a different story, because they're much more likely to re-sign Lee. Therefore, they should factor in the draft pick they saved by preemptively acquiring Lee.)
Further, Beane's adversaries have a July advantage that is increasing: bigger budgets. The AL West is very, very close, but the Angels and Rangers can (as demonstrated) toss more money at the situation than can the A's. Can the A's take on Alfonso Soriano's $3.5 million salary from here on out? Maybe, maybe not. Can the Angels? No doubt.
There's also less reason to think that the draft picks Beane would get from Soriano are as valuable as they were in, say, 2002. Why? Because teams are starting to follow Beane's draft strategy, grabbing known commodities such as established college players. Seventeen of the 30 first round selections in this year's draft were out of college, including the top six. Eight of the top 10 in 2005 were college kids. Compare this to 1999. That year, picks 3, 7, 9, and 10 were college students -- the others were high schoolers. At #9, the A's selected... Barry Zito. The Brewers grabbed Ben Sheets with the next pick. With college pitchers all the rage now, does anyone truly think either would fall that far? And if not, why is getting pick #19 or #33 or #47 so important?
Other general managers have followed Beane's lead both in "trading for draft picks" (via free agent coupons) and in their draft styles. There end result: Draft picks aren't worth as much as they used to be, and even if they were, the ability to get them has become more expensive.
The market inefficiency? It's an over-correction.
The way to exploit it? Do the opposite. In this case, it means shopping Zito now, before he hits free agency, with hopes that would-be buyers overestimate the compensatory issue. The Mets, for example,in formulating an offer, think that it's not just a Zito rental. Either (a) It's Zito, and we keep our #1 draft pick or (b) It's a Zito rental, plus we get two draft picks. If the Mets overvalue those draft picks, they'll offer more than they should. Beane can reap the benefit now -- and he can't do so later.
The Market's Future: The Flip
Throughout this article, I use the word "flip" to mean "to trade a recently-acquired player to another team." There's a reason for this -- if my analysis of the change of the market is correct, a good GM will look for a way to flip players often.
More specifically, the following should occur:
In November, the GM should identify players entering their walk year who are likely to be Class "A" free agents. He should try and acquire these players, if the price is right. (This includes current free agents that he can sign to a one-year contract, especially if they're not currently going to result in much or any compensation, but that's usually not the case.)
In calculating whether "the price is right," the GM should have an eye toward trading the player over the summer, absent, of course, a need to keep the player for a playoff run. Basically, the GM should use the player's "flip" value as a hedge against overpaying in November.
The old analysis would have the GM simply analyze whether the player, if added, would make the team better for the current season. But that analysis, as demonstrated above, is incomplete. The player does not lose value simply because the team fails to remain competitive. Rather, the player takes on a different value. He turns from on-field asset to trade-block commodity.
Because the incentive to acquire Class "A" free agents is high, the market for these players is increased. (That's why the Rangers could give up three MLB-ready players for two months of Carlos Lee and a good prospect; if they can't re-ink Lee, they'll get two draft picks!) And arguably, this incentive is overstated. Certainly, when you have multiple teams competing for limited commodities, the offering price will be on the high side.
The trick, again, is to acquire the July-tradable talent in November through January. Good teams or bad; it really doesn't matter. We'll see if that happens.
Date
Sun 07/30/06, 11:45 am EST
