I have been implementing an, in my opinion, interesting finite difference method (Runge-Kutta-Legendre of second order) to price American options in the standard Black-Scholes model (see "Pricing American Options with the Runge-Kutta-Legendre Finite Difference Scheme, F. Le Floc'h, International Journal of Theoretical and Applied Finance, vol 24, 2021"). This is an explicit method using so called super time-stepping.

My question is not related directly to the method itself though, but rather the choice of boundary conditions for the finite difference grid in the spatial direction. Many authors recommend assuming that the second derivative is zero on the boundary, $\frac{\partial^2 f}{\partial S^2}(t,S_{min}) = \frac{\partial^2 f}{\partial S^2}(t,S_{max}) = 0$.

However, this seems to lead to large errors in several cases, even if the upper boundary $S=S_{max}$ is chosen very large. The motivation given in articles and books seems to be that this condition is true for most payoffs.
While I agree that this is reasonable on maturity, the condition does not seem very appropriate when going backwards in time on the finite difference grid (Unless I made some trivial mistake, of course, which is certainly possible.).

To make things a bit more concrete, lets look at the (time-reversed) Black-Scholes equation: \begin{equation} \frac{\partial f}{\partial t} =\frac{1}{2} \sigma^2 S^2\frac{\partial^2f}{\partial S^2} + \mu S\frac{\partial f}{\partial S} - rf \end{equation} Here $r$ is the risk-free rate, $\mu = r - \delta$, where $\delta$ is the continuous dividend yield and $\sigma$ is the volatility.

Let us focus on the upper boundary. Assume we try to price a standard european put option with maturity in $5$ years, with strike $K=100$, where our upper spatial boundary on the grid is $S_{max} = 500$. For S = S_max, we then have \begin{equation} \frac{\partial f}{\partial t}(t,S_{max}) = \mu S\frac{\partial f}{\partial S}(t,S_{max}) - rf \end{equation} To make this clear, assume that $r=\mu=0$. Then our condition at the upper boundary becomes \begin{equation} \frac{\partial f}{\partial t}(t,S_{max}) = 0, \quad \forall t \end{equation} We have that \begin{equation} f(0, S_{max}) = max(K - S_{max}, 0) = 0. \end{equation} This together means that we will have $f(t, S_{max}) = 0$ for all t. The value gets stuck at 0. This results in large errors, especially for longer maturities and higher volatilities. In fact, we can see that the second order term we assumed to be zero is actually far from zero, especially compared to the other terms in the Black-Scholes PDE (which were actually zero in our case).
More precisely If we look at the second order term at the time we price the option (real non reversed time $t=0$) \begin{equation} \frac{1}{2} \sigma^2 S^2\frac{\partial^2f}{\partial S^2}, \end{equation} we have that this is $\frac{1}{2} \sigma^2 S_{max}^2 \cdot \Gamma$, where $\Gamma$ is the gamma of the option. We can calculate this explicitly for european options. We have \begin{equation} \frac{1}{2} \sigma^2 S_{max}^2 \cdot \Gamma = \frac{\sigma \cdot S_{max}}{2 \sqrt{T}} e^{-d_1^2/2}, \end{equation} where \begin{equation} d_1 = \frac{\log(\frac{S_{max}}{K}) + T \cdot \frac{\sigma^2}{2}}{\sigma \cdot \sqrt{T}} \end{equation} This is not close to zero unless $S_{max}$ is extremely large, so ignoring this term typically results in a substantial error, if I did not make a mistake.

So the question is if I am doing something wrong in my reasoning here. I can see the errors clearly on plots and also the theory seems to confirm that. But since the condition is so commonly recommended I wonder if I might have missed something here?

Is there any easy solution for this problem? One way I found around this is to make lots of points with distance between them growing exponentially outside the original upper boundary. But this makes calculations take more time as well.

  • $\begingroup$ Could you please write what you mean with "extremely" large? E.g. would a level of $S_{max}=500$ be extreme when $S_0=100$? $\endgroup$ May 9, 2022 at 11:50
  • $\begingroup$ I would not call $S_{max} = 500$ especially large here. This choice makes the mentioned second order term, $\frac{1}{2} \sigma^2 S_{max}^2 \cdot \Gamma$, using the values $\sigma = 0.7$, $T = 5$ equal to around 15. If $S_{max} = 5000$, we get a value of around 3.6. Still not very close to 0. $\endgroup$ May 9, 2022 at 12:18
  • $\begingroup$ The problem becomes very visible on plots if the volatility and the time to maturity is increased (which seems reasonable, since we then make a larger error by assuming the diffusion term to be zero. Also the probability of hitting the upper boundary, so that the boundary condition comes into play, increases if the time to maturity is increased.) $\endgroup$ May 9, 2022 at 12:26
  • 1
    $\begingroup$ The time reversed Black-Scholes equation starting at maturity going backward (which is what we want, since we have our initial conditions at maturity) is the stable way to go. This time reversed Black-Scholes is essentially equaivalent with the heat equation going forward in time. Going forward in time using the Black-Scholes equation would make the finite-difference scheme unstable. $\endgroup$ May 9, 2022 at 13:37
  • 1
    $\begingroup$ The BS equation you typed into your question has (if I am not mistaken) the sign of the $\partial t$ term flipped. $\endgroup$
    – Kurt G.
    May 9, 2022 at 13:56

1 Answer 1


Alright, so I solved my issues here. Since it might be useful to others stuck on the same thing, I mention it here. The problem I described about assuming the second order term was equal to zero is real, but as I mentioned, the problem goes away if the upper limit of the grid is huge.

My solution was to not use a uniform grid, but rather to create a uniform grid on some interval $[-C,C]$ for some large enough $C$, say $C=15$. We get grid points $x_i = -C + dx \cdot i, \quad i= 0, 1,2, \ldots$ for $dx = 2C/ (\textrm{number of spatial steps})$.
Then I create the grid points in the asset variable as $S_i = e^{x_i}$ for each gridpoint $x_i$ on the uniform grid. In this way one can reach the endpoint in the upper grid far way (where the second order term is negligible) without using too many steps in the spatial direction.

This is of course not some new magic trick, but rather imitates how the grid would transform after making a change of variable to the log space $x = log(S)$, which is one of the most common way to transform the Black-Scholes equation before doing any numerical calculations. However, I wanted to see if I could do everyting in the original formulation without using this transform. So it turned out fine, after using the "exponential" grid described. The problem was due to me using a uniform grid.

  • 2
    $\begingroup$ I think this is the right answer. People often use 5 standard deviations as the grid boundary and work in log spot space, and I always found good convergence with that. You might also be interested in this paper I wrote, which shows how to discretize and set b.c.s so that vanillas on the grid are exactly recovered papers.ssrn.com/sol3/papers.cfm?abstract_id=3530561 $\endgroup$
    – Peter A
    May 10, 2022 at 9:37
  • 1
    $\begingroup$ Thanks, I will have a look at your paper. $\endgroup$ May 10, 2022 at 10:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.