Aharanov-Bohm Type Arbitrage and Homological Obstructions in Financial Markets

This paper introduces a novel concept of Aharonov-Bohm-type arbitrage in financial markets, demonstrating how global topological inconsistencies (non-trivial holonomy) in filtered systems can generate profitable, predictable trading strategies despite the absence of local price discrepancies.

The Gist

The Big Idea: The Invisible Loop Profit

Imagine you are walking through a city. Usually, if you walk in a circle and end up exactly where you started, you expect to have the same amount of money in your pocket as when you left. If you spent \5 on a coffee and earned \5 on a lemonade stand along the way, you are back to square one.

In traditional finance, "arbitrage" (making risk-free profit) is like finding a store that sells a shoe for \10 and a store down the street that buys it for \12. You see the price difference locally and make a quick profit.

This paper proposes a weird new kind of profit. It suggests that sometimes, you can walk a perfect circle, do everything "fairly" at every single step, and yet, when you return to your starting point, you have more money than when you left.

This is called Aharonov–Bohm (AB) Arbitrage. It's named after a famous physics experiment where a particle gains energy just by traveling in a loop around a magnetic field, even if it never touches the field itself. In this financial version, the "magnetic field" is a hidden, global inconsistency in the market that you can only see by completing a full loop.

---

The Core Concepts, Explained Simply

1. The "Distortion" (The Leaky Bucket)

Imagine you are passing a bucket of water (your money/information) from person to person in a line.

• Normal Market: If Person A passes the bucket to Person B, and the rules say "keep the water level the same," the bucket arrives full.
• This Paper's Market: Sometimes, the rules of the game change slightly as you move. Maybe Person A pours a little extra water in, or Person B accidentally spills a little. The paper calls this a "Distortion."

If you look at just one step (A to B), it might look like a tiny, harmless change. But the paper argues that these tiny changes don't always cancel out.

2. The Loop (The Round Trip)

Now, imagine a group of traders passing that bucket in a circle:

• Trader A $\to$ Trader B $\to$ Trader C $\to$ Trader A.

In a normal world, if you pass the bucket around the circle, it should be the same size when it gets back to A.
But in this "distorted" market, the bucket might arrive bigger.

• Maybe A gave B a full bucket.
• B gave C a slightly larger bucket (because of a hidden rule).
• C gave A an even larger bucket.

When the bucket returns to A, it's overflowing. A has made a profit without ever doing anything "illegal" or "unfair" at any single step. The profit comes from the shape of the loop, not the individual steps.

3. The "Holonomy" (The Magic Number)

The paper uses a fancy math word called Holonomy. Think of it as a "Magic Number" that tells you how much the bucket grew after the full trip.

• If the Magic Number is 1, the bucket is the same size. No profit.
• If the Magic Number is 1.1, the bucket grew by 10%. Profit!

The paper shows that you can calculate this Magic Number just by looking at the rules of the market at the very beginning, before you even start the loop.

4. The Trading Strategy (The Bet)

How do you actually make money from this?
The paper suggests a simple strategy:

1. Check the Magic Number: Before you start, calculate the "Holonomy" of a specific loop of trades.
2. The Bet:
   • If the Magic Number says you will end up with more money, you execute the loop (buy low, sell high, buy low, sell high in a circle).
   • If the Magic Number says you will end up with less money, you do the reverse loop (sell high, buy low, etc.).
   • If it's neutral, you do nothing.

Because the Magic Number is calculated using only information available right now, this isn't "cheating" or predicting the future. It's just spotting a hidden geometric flaw in the market's structure.

---

A Real-World Analogy: The Currency Exchange Maze

Imagine a currency exchange office in a mall with three booths: Dollars, Euros, and Yen.

• Booth 1 (Dollars to Euros): They say, "Give me $1, I'll give you €0.90." (Normal).
• Booth 2 (Euros to Yen): They say, "Give me €1, I'll give you ¥100." (Normal).
• Booth 3 (Yen to Dollars): Here is the twist. They have a hidden rule: "If you come from the Euro booth, we give you a bonus." So, if you give them ¥100, they give you **\1.05** instead of the usual \1.00.

The Local View:
If you stand at Booth 1, you think, "Okay, $1 becomes €0.90."
If you stand at Booth 2, you think, "Okay, €1 becomes ¥100."
If you stand at Booth 3, you think, "Okay, ¥100 becomes $1.05."
Everything looks fine locally. No one is cheating.

The Global Loop:
You start with $1.00.

1. Go to Booth 1: You get €0.90.
2. Go to Booth 2: You get ¥90.
3. Go to Booth 3: You get $0.945 (Wait, that's a loss? Let's adjust the numbers to match the paper's logic).

Let's try a better example where the "Distortion" accumulates:
Imagine the exchange rates are slightly different depending on who you are or where you came from, but the signs don't say that.

• Step 1: $1 $\to$ €1.02 (Hidden bonus).
• Step 2: €1.02 $\to$ ¥102 (Hidden bonus).
• Step 3: ¥102 $\to$ $1.05 (Hidden bonus).

You started with \1. You ended with \1.05.
You walked a circle. You didn't break any rules at any single booth. But the system as a whole allowed you to print money.

The paper says: "If you can find these loops, you can build a machine that prints money."

---

Why Does This Matter?

1. It's Not Just Math: The paper proves that this isn't just a theoretical curiosity. If these "loops" exist, you can actually trade them to make money.
2. It's Global, Not Local: Traditional regulators look for bad apples (one person cheating). This paper says, "Look at the whole system. Even if everyone is honest, the shape of the market might be broken."
3. The "Admissibility" Check: The paper admits that just because a loop mathematically works, doesn't mean you can physically do it. Maybe the transaction fees are too high, or you can't trade fast enough. They call this "Admissibility." If a loop is "Admissible," it's a real money-maker. If not, it's just a math puzzle.

Summary in One Sentence

This paper discovers that in complex financial markets, you can sometimes make risk-free profit not by finding a cheap item, but by walking a specific circle of trades where the market's hidden rules accidentally add up to a bonus at the end of the day.

Technical Summary

1. Problem Statement

Traditional arbitrage theory relies on local consistency conditions. In discrete-time models, the absence of arbitrage is characterized by the existence of equivalent martingale measures, ensuring that price increments have zero conditional expectation at every individual time step. These formulations are inherently local, constraining transitions between adjacent time points.

The paper identifies a gap in this framework: it does not account for global inconsistencies that may arise from the cumulative effect of transitions along a closed loop, even if every individual transition appears locally neutral (i.e., no local arbitrage exists). The author seeks to formalize a new type of arbitrage arising from global loop effects, analogous to the Aharonov–Bohm (AB) effect in quantum physics, where a particle acquires a phase shift along a closed path despite the absence of a local field.

2. Methodology

The paper employs category theory and homological algebra to model financial markets and information flows.

A. Categorical Setup

• Time/Information Structure: Modeled as a small category $\mathcal{T}$.
• Filtration: Defined as a contravariant functor $F: \mathcal{T}^{op} \to \mathbf{Prob}$, where $\mathbf{Prob}$ is the category of probability spaces with null-preserving morphisms.
• Conditional Expectation: A functor $E: \mathbf{Prob} \to \mathbf{Ban}$ maps probability spaces to their $L^1$ spaces and morphisms to conditional expectation operators.
• Composed Functor: The system is analyzed via the composition $E \circ F: \mathcal{T}^{op} \to \mathbf{Ban}$.

B. The Distortion Function ($d_F$)

A key observation is that the conditional expectation operator $(E \circ F)(i)$ does not generally preserve constant functions (unlike in measure-preserving systems). This leads to the definition of a canonical multiplicative distortion:
$$d_F(i) := (E \circ F)(i)(1_{F(t)}) \in L^1(F(s))$$
for an arrow $i: s \to t$.

• If the transition is measure-preserving, $d_F(i) = 1$.
• If not, $d_F(i)$ measures the deviation, acting as a multiplicative vector potential.

C. Generalized Martingales

The paper redefines the martingale condition to account for this distortion. A family $f = (f_t)$ is an $F$-martingale if:
$$(E \circ F)(i)(f_t) = f_s \cdot d_F(i)$$
This generalizes the classical condition (where $d_F(i)=1$).

D. Holonomy and Loop Effects

For a closed loop $\gamma = (i_1, i_2, \dots, i_n)$ in $\mathcal{T}$, the holonomy $Hol(\gamma)$ is defined by inductively transporting and multiplying distortions along the path. This represents the cumulative global distortion.

• Multiplicativity: The distortion satisfies a cocycle-like relation: $d_F(j \circ i) = (E \circ F)(i)(d_F(j))$.
• Holonomy Calculation: $Hol(\gamma)$ is computed by transporting the distortion factors back to the base time $t_0$.

3. Key Contributions

1. Definition of AB Arbitrage:
   The paper introduces Aharonov–Bohm (AB) arbitrage, defined as the existence of a loop $\gamma$ where the holonomy $Hol(\gamma)$ satisfies:
   $$\mu_{t_0}(Hol(\gamma) > 1) > 0$$
   This represents a global inconsistency invisible at the level of individual transitions but detectable upon completing a round-trip trade.

2. Cohomological Interpretation:
   The distortion $d_F$ is identified as a multiplicative 1-cocycle. The paper establishes a conceptual link between cohomological structures (specifically the triviality of global invariants/holonomy) and the economic concept of "no-arbitrage."

3. Translation to Trading Strategies:
   The author demonstrates that non-trivial holonomy is not merely a mathematical artifact but can be converted into a predictable, self-financing trading strategy.

   • Strategy: At time $t_0$, observe $Hol(\gamma)$. If $Hol(\gamma) > 1 + \epsilon$, execute the loop $\gamma$. If $Hol(\gamma) < 1 - \epsilon$, execute the reverse loop $\gamma^{-1}$.
   • Predictability: The decision is based solely on information available at $t_0$ (measurable with respect to $F(t_0)$), ensuring the strategy is non-anticipative.

4. Admissibility Conditions:
   The paper formalizes admissibility to distinguish between theoretical loop effects and economically realizable strategies. An admissible loop must satisfy:

   • Observability (measurability of holonomy).
   • Executability (availability of market transactions).
   • Composability (sequential execution).
   • Self-financing (no external capital injection).
   • Reverse executability.

4. Results

• Proposition 1: The distortion satisfies a multiplicative consistency relation, behaving as a transport mechanism.
• Proposition 2: If all transitions are measure-preserving ($d_F = 1$), the holonomy is trivial ($Hol(\gamma) = 1$), and no AB arbitrage exists.
• Illustrative Example: A concrete example with three time steps ($t_0, t_1, t_2$) and non-measure-preserving transitions (involving information loss and re-injection) is constructed. The calculation shows $Hol(\gamma) = 4 \cdot 1_{\{1\}}$, yielding a probability of strict gain of $0.5$, thus proving the existence of AB arbitrage in this system.
• Proposition 3 & 4: It is proven that if an admissible loop has non-trivial holonomy, the defined threshold strategy yields a terminal wealth $V_{out} \geq 1$ with a positive probability of strict gain ($V_{out} > 1$).

5. Significance

• Paradigm Shift: The paper shifts the perspective of arbitrage from local (step-by-step consistency) to global (loop consistency). It suggests that markets can be locally efficient (no local arbitrage) yet globally inefficient (AB arbitrage exists).
• Interdisciplinary Bridge: It successfully bridges categorical methods (functors, natural transformations, cohomology) with financial mathematics (martingales, arbitrage, trading strategies).
• New No-Arbitrage Condition: It proposes that the absence of arbitrage should be characterized by the triviality of global invariants (holonomy) rather than just the existence of local martingale measures.
• Future Research Directions: The framework opens avenues for:
  • Developing a precise cohomological theory for filtered market systems.
  • Incorporating market frictions (transaction costs) into the admissibility conditions.
  • Re-evaluating the Fundamental Theorem of Asset Pricing through the lens of global topological obstructions.

In summary, Adachi's work provides a rigorous mathematical framework where arbitrage is interpreted as a topological obstruction (non-trivial holonomy) in the space of market transitions, offering a novel tool to detect and exploit global market inconsistencies that evade traditional local analysis.