[ge_arrow] Correct summation axis

[UdohNakamura]

Hello,

Thank you for creating and maintaining this excellent lecture note. It has been very helpful for my studies.

I noticed a potential inconsistency between the theoretical definition of the risk-free rate and its implementation in the RecurCompetitive class.

The Issue:

• In the "Markov Asset Prices" section, the price of a risk-free bond in state i is defined as $R_i⁻¹ = Σ_j Q_{ij}$. This corresponds to summing the Q matrix across its rows (axis=1 in NumPy).

• The current implementation in the risk_free_bond function uses np.sum(self.Q, 0), which sums across columns (axis=0).

The Fix:

This PR corrects the summation axis from 0 to 1 to align the code with the theoretical text.

Please let me know if I have misunderstood anything.