one development card, please!
It just so happens that Catan resembles the model of liquidity shocks in Michael Pettis's Volatility Machine―and winning Catan calls for strategies that approximate his advice.
image: a development card!
***
April 3, 2024
I was playing Settlers of Catan a few weeks ago with my housemates and, man, I was losing. My early-game strategy of building cities had burnt itself out as the dice rolls just refused to give me resources I needed to continue growing my territory. I had no development cards to change any odds in my favor, all while my rivals started encroaching on everything I had built. If only the dice were in my favor!
This is a bad way to play Catan. Unfortunately, this is also what usually happens to me when I play. Come mid-game, I’m always frustrated that I haven’t hedged my bets adequately. I’m relying on too few tiles for resources, and I haven’t bought enough development cards. None of this is a problem when the dice are on my side. But, more often, when they aren’t, I can’t support myself.
I was reading Michael Pettis’s Volatility Machine that same week and, as I realized I had no shot of winning this game, I thought, “hold on a minute, Catan is kind of like the Volatility Machine!”1
The Volatility Machine, from 2001, makes the case that the financial crises that emerging markets faced in the 80s and 90s were the result of their governments’ irresponsible risk management decisions: policymakers failed to insulate their sovereign debt and their currencies against exogenous liquidity shocks. Better economic policymaking, Pettis argues, requires governments to take a page out of the corporate finance risk management textbook and hedge their bets.
Now, it just so happens that Catan resembles the model of liquidity shocks that Pettis introduces in the Volatility Machine―and winning Catan calls for strategies that approximate Pettis’s advice.
To win Catan, players need to reach ten victory points, earned through building cities, settlements, roads, and armies, all paid for by resources earned when the dice rolls are favorable. Players set up a board of resource tiles and place settlements in the best possible locations, contingent on each others’ settlement placements. A die is cast, players earn resources depending on which of their settlements touch resource tiles that match the result of the dice roll, and players can trade their resources to purchase new roads, settlements, and cities to expand their ability to secure resources on future dice rolls. There are additional dynamics at play, such as ports to facilitate trade, robbers to randomly enable resource theft, and development cards, which I’ll describe soon, but this simplified depiction should suffice.
Every Catan player has their assets and their liabilities. Your assets include your short-term ability to grow your victory points and your victory point earning potential by building new settlements, cities, and roads with the resources you possess. Your liabilities include your short-term exposure to being robbed, being denied access to a resource, or being cut off from territory that could earn you victory points. Essentially, the value of your liabilities is the value of the victory points you expose yourself to failing to secure. This is all very inexact, I know―Catan is not high finance no matter how hard I try to place futures contracts―but bear with me.
Catan players are at the mercy of the dice. This is our exogenous liquidity variable. Sometimes it’s good, and we’re rolling in the dough (wheat), but, more often, it’s trash―our rivals get all the good rolls, we’re being stolen from, or there’s just a persistent resource shortage that prevents you from developing.
By default, you can ride the randomness wave and hope for good rolls and good trades. But, if it’s just too frustrating to be a persistent victim of bad dice rolls, you can buy insurance: for the admittedly steep price of an ore, a sheep, and a wheat, you can purchase development cards!
Each type of development card enables you to hedge against bad times―liquidity crunches―in a different way. Knights are like robbers, allowing you to deny other players from earning certain resources even if the dice are in their favor, and activating three of them earns you the largest army award, worth two victory points; monopoly cards allow you to take every unit of a certain resource from every other player’s hand; years of plenty allow you to claim two resources; road builder builds you two roads; and certain development cards are hidden victory points, lowering your win threshold. Of course, these insurance strategies themselves are unevenly distributed through the deck of development cards. You’re always more likely to get a knight than a victory point. But what matters is that development cards empower you to continue claiming resources and warding off your rivals, especially when neither the dice rolls support you nor your rivals have any interest in trading with you. Crucially, development cards can’t be stolen or taken away.
I think these two strategies―trusting the dice and buying insurance, respectively―are analogous to Pettis’s conceptions of inverted and correlated capital structures. In an inverted capital structure, the values of market participants’ assets and liabilities are inversely correlated: when asset values rise, liability values fall, and vice versa. In a correlated capital structure, or a hedged capital structure, the values of assets and liabilities are instead positively correlated: asset and liability values rise and fall in tandem. These capital structures are safer and less volatile because market participants’ payment obligations fall when they have less ability to pay and rise when they have more ability to pay.
Playing Catan, I often find myself with what amounts to an inverted capital structure. I end up relying on a few resource tiles to build cities in the hopes that they can consistently generate the bulk of the future resources and trading opportunities I need to expand my territory and earn victory points. If the dice rolls are right, I rightfully earn the ire of my rivals. But once the dice rolls stop consistently landing on the concentrated set of numbers that benefit me, my mass industrialization program stops in its tracks. An exogenous liquidity shock has tanked my victory point earning potential, and my liabilities skyrocket in comparison. My fields lie fallow while my rivals eat my freaking lunch.
Every time I lose a game of Catan, which is too often, I realize I should have invested in a more correlated capital structure. My hubris is to go for broke building cities rather than buying insurance: I don’t often stock up on development cards to lay contingent claims on others’ resources when I might need them most. Hedging my capital structure with knights and monopolies, for example, would allow me to ward off against my rivals spending resources I’d want to seize for myself. And nobody could steal my development cards with their knights or robbers. A few bad dice rolls may still significantly lower my victory point earning potential, but at least I keep my liabilities in check.
Development cards are not the only parts of a good correlated capital structure. It’s important to build your roads, cities, and settlements in such a way that you don’t depend on a few dice rolls. The probability distribution of the sum of two even six-sided dice is bunched up around 6, 7, and 9, meaning players should build around 5, 6, 8, and 9 where possible (in Catan, a 7 activates the robber). But, in any given game, the actual distribution of rolls is hardly evenly distributed. The first half of the game might see more 6s rolled than 8s, for example, annoying everyone with access to 8s while making the 6s feel giddy with purchasing power. Nobody said liquidity would be distributed fairly, if at all. So it’s important to build territory in such a way where most dice rolls give you at least something, if not nothing.
I know this, but I don’t often take these lessons to heart. In the past month, I’ve played Catan three times. The first time, I lost squarely because my capital structure was simply just too inverted. This loss inspired this post. The second time, my capital structure looked a bit better, but I was just barely defeated when, really at the very last minute, my rivals ganged up on me to deny me resources and throw the victory to my housemate. (Quite unfair, although this happens in the real global economy, too.) The third time, I snatched victory from the jaws of defeat: my resource distribution was way too concentrated around too few possible dice rolls and I didn’t buy enough development cards, but a few rounds of bad dice rolls and robbers kept any of my rivals from clinching a victory for just as long as I needed to get lucky. It was really close, though!
It’s hard to know beforehand if your capital structure is too inverted or not correlated enough. I only realize I need insurance when I can’t do anything else. Pettis’s framework is similar: when the liquidity is flowing, it’s legitimately tough for emerging market leaders to distinguish legitimate policy-driven growth from speculative investment. Perhaps this makes a good game of Catan—I end every game wondering if I could have played more tactfully or if I was merely the victim of bad dice rolls—but Pettis certainly doesn’t think this dynamic helps the global economy. We don’t have the luxury of making bad choices! Development is at stake.
***
Catan is premised on settling land and trading commodities. It’s very nakedly colonialist. It’s also extractivist, but the base game doesn’t build in any ecological or climatic consequences for resource extraction.
Recognizing this, the Catan game developers actually made an oil extraction expansion pack for the game, with climate impacts included, to boot.2 I haven’t played it, but I’m sure it would be interesting to try. And, just this week, a whole new edition of Catan called New Energies was announced, too. The new game apparently forces players to consider pollution, fossil fuels, clean energy, and, also, suspiciously, overpopulation as they pursue victory points. In short, New Energies seems to have lots of the same dynamics as the earlier oil extraction expansion. Both expansions add extra volatility to the game in the form of variable resource yields and natural disasters, the quotidian consequences of climate change. But this volatility is endogenous to the choices players make about resource extraction, reminiscent of the climate risk “doom loop,” rather than the purely exogenous volatility of a dice roll that doesn’t go your way. Now players have to make choices with the knowledge that every step they take toward winning could end up consigning the whole table to catastrophe.
I suspect that all these rules make Catan way more complex than it probably deserves to be. But I’ll give their designers credit for their attempts at building realistic endogenous volatility dynamics into the game.
What would the Volatility Machine say about the oil expansion? What are Pettis’s guidelines for how market participants should hedge against shocks, and how do they square up against the endogenous volatility of climate change?
Per Pettis’s model, the liquidity shocks that emerging markets face happen for their own reasons3 in the Global North. Emerging markets receive these shocks as exogenous volatility in capital flows. I think this judgment holds up. Global North liquidity conditions are pretty exogenous to Global South economic policy, if the disastrous impacts of the Fed’s hiking cycle on Global South debt sustainability are any indication.
To ward against these shocks, Pettis offers four guidelines. First, flexibility: shocks shouldn’t threaten market participants’ ability to raise cash and meet contractual obligations. Second, stability: market participants must be able to protect themselves in such a way that changes in external market conditions cannot cause them to abandon their existing investment and growth strategies. Third, volatility minimization: correlated capital structures! Fourth, relative value: don’t miss out on good deals and arbitrage opportunities that investors and trading partners might be interested in―insofar as those speculative opportunities don’t threaten the achievement of the other three guidelines.
I’ve been chewing on these guidelines since I finished the book and, to be honest, I’m not sure they really work for endogenous shocks. Certain market participants’ investment and growth strategies―in the real world, and in Catan―rely on the kinds of resource extraction that call forth increasingly frequent and increasingly deadly climate shocks. Neither stability nor flexibility is guaranteed, insofar as market participants’ growth strategies cut into each others’. Those more exposed to shocks will have a harder time raising new cash. And volatility minimization gets harder and harder to achieve as hedging and insurance grow more expensive. (In a way, this just rehashes my doom loop argument.) Finally, pursuing “relative value” is always possible, but Pettis (and I) don’t place much faith in speculation as a source of long-term growth or stability.
The Volatility Machine is set on reinterpreting how individual emerging markets should play the global macroeconomic field, so to speak. Where climate change is concerned, however, playing the field successfully has dubious outcomes. Individual countries’ development strategies might insulate them from exogenous market fluctuations today, and even provide them with short-term gains, but those same strategies will slowly but surely hurt them and other governments in the coming years in the form of endogenous shocks. I’m reminded of how Indonesia uses coal plants to power nickel smelters, a core plank of its industrialization strategy, and of how Guyana is growing rapidly on the back of Exxon’s oil prospecting. It doesn’t really matter if Indonesia’s or Guyana’s capital structures are inverted or correlated or whatever―the implication here is that the development trajectories taken by individual countries can expose all countries to greater climate vulnerabilities.
Granting that certain countries are far more liable for the damages than others are, it’s clear that good climate policymaking requires that no government play the field for itself. Economic coordination allows governments and other market participants to hedge against shocks better than individualized liability management ever could. Applying Pettis’s guidelines across all market participants helps illuminate what the goals of global climate policy coordination should be: to ensure that no market participants’ growth strategies exacerbate climate change, then to ensure that climate shocks don’t threaten market participants’ ability to raise funds or pay creditors, and, finally, to ensure that market participants’ capital structures remain correlated such that they do not need to worry about paying creditors when they cannot raise funds. All this coordination requires a global liquidity backstop.
But global climate policymaking, at least where the financial system is concerned, advances none of these goals, at least not in any systematic way. That’s too bad: for the Volatility Machine’s guidelines to actually be useful for addressing the climate challenge, we have to treat the whole world as one corporation―a corporation that remains, in perpetuity, a wholly owned subsidiary of nature. Until then, the guidelines will become less and less useful for individual market participants as their growth strategies cut against one another's.
It’s the same for the oil expansion version of Catan, where it’s possible to win not necessarily by securing enough victory points, but through what’s labeled a pyrrhic victory: “If the fifth number token is removed from one of the hexes, flooding has overwhelmed Catan and all inhabitants are forced to abandon the island, thus ending the game. While no player truly wins, the player who currently holds the Champion of the Environment token achieves a ‘Pyrrhic Victory.’ That player is recognized by the international community for [their] efforts to mitigate climate change and is granted the most attractive land on a neighboring island to resettle.”
This is where all my fun parallelisms fall apart. There is no neighboring island to resettle. And, if and where people can resettle, doing so comes at the cost of irreversible losses. All we have is our one blue marble, and nobody is going to care who the most “green” individual was during a climate catastrophe, because they clearly weren’t green enough. Playing Catan, the default version or its climate-conscious expansions, is exhilarating―but it’s certainly no model for how to play the economy, no matter what financial principles I think it illustrates. There’s no “winning” unless everyone wins together, and we can’t play the economy again next week.
footnotes
Guy who’s only read Volatility Machine: “getting a lot of Volatility Machine vibes from this game.”
The National Oceanic and Atmospheric Administration (NOAA) also showcases an educational “global warming” expansion to the game, intended for students.
The reason is Minsky. Move along.