#pragma GCC optimize("Ofast")
#include <bits/stdc++.h>

using namespace std;

typedef long long llint;

const int MAXN = 5000;

llint a[MAXN], b[MAXN], k[MAXN], k2[MAXN];
llint dp[MAXN + 1][MAXN + 1];

int main()
{
	ios_base::sync_with_stdio(false);
	cin.tie(nullptr);
	
	freopen("fruits.in", "r", stdin);
	freopen("fruits.out", "w", stdout);
	
	int n, i, j;
	
	cin>>n;
	for (i = 0; i < n; ++i)
		cin>>a[i];
	for (i = 0; i < n; ++i)
		cin>>b[i];
	for (i = 0; i < n; ++i)
	{
		cin>>k[i];
		--k[i];
		k2[i] = k[i];
	}
	
	sort(k2, k2 + n);
	int m = unique(k2, k2 + n) - k2;
	for (i = 0; i < n; ++i)
		k[i] = lower_bound(k2, k2 + m, k[i]) - k2;
	
	for (i = n - 1; i >= 0; --i)
	{
		memcpy(dp[i], dp[i + 1], sizeof dp[i]);
		for (j = k[i]; j >= 0; --j)
		{
			dp[i][j] += a[i] + k2[j] * b[i];
			if (dp[i][j + 1] > dp[i][j])
				dp[i][j] = dp[i][j + 1];
		}
	}
	
	cout<<dp[0][0]<<'\n';
}