Trading and Exchanges Chapter 17: When Arbitrage Goes Wrong (Part 2)
In Part 1 we covered what arbitrage is, the different types (pure vs speculative), and how arbitrageurs keep prices consistent across markets. Sounds like easy money, right? Buy low here, sell high there, pocket the difference. This part is about why it is not that simple. Harris lays out four specific risks that make arbitrage genuinely dangerous, and he has some incredible real-world examples to prove it.
The Four Risks of Arbitrage
Harris identifies four risks that can ruin an arbitrageur’s day. Or career.
Implementation risk is the risk that you cannot execute your trades at the prices you expect. Arbitrage requires trading multiple instruments at the same time. If you buy one leg but the other leg moves before you can trade it, your carefully calculated profit evaporates. Harris compares this to jumping between ice floes. When both feet are on the same floe, you are fine. When you are straddling two floes that are drifting apart, you are about to get very wet.
Arbitrageurs manage this by trading all legs simultaneously, or alternating fills between instruments in illiquid markets. It is like loading a small boat. Put all the weight on one side first and you capsize.
Basis risk is the risk that the price gap you are betting on moves against you instead of closing. For pure arbitrages, the basis must eventually converge. But “eventually” can be a long time, and you might run out of money before it happens. Bigger portfolios mean bigger risk. Harris is very clear: never leverage to the maximum your capital allows. You need staying power. If the basis widens and you cannot fund your losses, your broker liquidates your position at the worst possible moment.
Model risk is the risk that you misunderstood the relationship between the instruments in the first place. Maybe what you thought was an arbitrage opportunity was actually the basis moving toward its fair value, not away from it. You end up on the wrong side of the trade.
Harris tells a fantastic story here about Fischer Black at Goldman Sachs. In 1984, Black figured out that the ValueLine Index futures contract was mispriced because other traders used an approximate formula instead of the exact one. The exact formula depended on individual stock volatilities, which everyone else ignored. Goldman traded aggressively on the correct formula. They made over $125 million. The other arbitrageurs, trading on the wrong model, got destroyed. The Kansas City Board of Trade eventually lost its entire market share in stock index futures because of this. One professor with the right math versus an entire market with the wrong math.
Carrying cost risk is the risk that holding your position costs more than you expected. Interest rates might rise, dividends might not come through, short positions might need unexpected funding. If prices quadruple while you are holding a short position, you need to come up with the cash to cover the margin. Harris shows a real example with Mechanical Technology and Plug Power where an arbitrageur needed $72,640 in additional funds after prices quadrupled in two months. If you cannot pay, your broker closes you out.
The LTCM Disaster
The most famous example of arbitrage risk is Long-Term Capital Management. It illustrates every risk at once.
LTCM was a hedge fund that did pure and near-pure arbitrages in bond and swap markets. They used massive leverage. On July 31, 1998, they controlled $125 billion in assets on $4.1 billion of equity capital. Their futures positions exceeded $500 billion. Their swaps exceeded $750 billion.
Then Russia defaulted on its bonds in August 1998. Credit and liquidity spreads widened dramatically. LTCM’s positions lost value. But here is the painful part: most of their positions actually became more attractive than ever. The basis widened, meaning the expected profit was even larger.
Did not matter. By September 21, LTCM was days from default. A consortium of creditors injected $3.6 billion for 90% of the firm. The partners lost nearly everything.
The positions eventually recovered. Had LTCM survived the crisis, the partners would have made fortunes. They were right about the arbitrage. They were wrong about how much capital they needed to survive until the arbitrage paid off. Classic lack of staying power.
Why Arbitrage Opportunities Exist
Harris explains two scenarios that create arbitrage opportunities.
First, the slow price adjustment scenario. New information arrives but not all prices adjust at the same speed. One market reacts instantly, another lags. Arbitrageurs jump in and push the lagging prices toward correct values. In this case, speed matters. You need to trade fast before the opportunity disappears.
Second, the uninformed liquidity demand scenario. No new information, but heavy buying or selling by uninformed traders pushes prices away from fair value in one market. Arbitrageurs step in and supply liquidity, pushing prices back. Here, you can be more patient.
Arbitrageurs, Dealers, and Brokers
Harris makes a clever observation. Dealers supply liquidity across time, matching buyers and sellers who show up at different moments in the same market. Arbitrageurs supply liquidity across space, matching buyers and sellers who show up at the same time in different markets. They compete with each other, and neither loves the other for it. Arbitrageurs also function like brokers, matching buyers with sellers across markets, except brokers earn commissions while arbitrageurs earn the spread.
The Bottom Line
Competition among arbitrageurs narrows spreads until only the most efficient survive. In markets where the math is easy, the fastest traders win. In markets where the math is hard, the best analysts win.
The biggest takeaway from this chapter: “risk-free profit” is a textbook fantasy. Real arbitrage involves execution risk, model risk, carrying costs, and the terrifying possibility that you are right about the trade but run out of money before it pays off. Just ask the partners at LTCM.
Previous: Chapter 17: Arbitrageurs (Part 1)