Computing the accrued interest in trinomial tree

user_17833660 0
Viewed 34


I'm currently trying to compute the accrued interest in a trinomial Hull-White tree.

The setup is a bond with a facevalue of 100 that matures in 6 years with a coupon of 5% (annually compounding).
The tree has 10 time steps, meaning that dt = 0.6. For each node time, i.e. 0,0.6,1.2,1.6 etc. I made a loop, that asks if the node-time has reached a coupon date, i.e. 0,1,2,3.. etc.

If a coupon is reached (e.g the coupon date 1 is reached at node-time 1.2 and 1.6) the accrued should be (1.2-1)*5=1 and (1.6-1)*5=4 respectively. I want to end up with a accrued at each node from 0 to 9 of [0,3,1,4,2,0,3,1,4,2]

But what i get now is [0,3,1,9,2,0,3,1,24,2].

My code is:

´´´

timesteps = 10
maturity = 6
coupon_dates = np.full(maturity+1, range(maturity+1))
dt = maturity/timesteps
node_times = [i * dt for i in range(timesteps)]
accrued = np.zeros(timesteps)

  for i in range(len(node_times)):
        accrued[i] = round(node_times[i] * 5)
        for j in range(len(coupon_dates)):
            if (node_times[i]-dt < coupon_dates[j] and node_times[i] >= coupon_dates[j]):
                accrued[i] = round((node_times[i] - coupon_dates[j]) * 5)

´´´

Can anyone help me with this problem? And do someone have a more efficient way to compute the accrued?

Thanks in advance!

for-loop python
1 Answers

I solved the problem myself.

´´´

    coupon_dates = [0,1,2,3,4,5,6]
    timesteps = 10
    accrued = []
    dt = 6/10
    coupon = 0.05
    face_value = 100
    for t in range(timesteps):
        coupon_date = max(i for i in coupon_dates if i <= t * dt)
        accrued.append((t * dt - coupon_date) * coupon * 100)

´´´

user_178336600
Related