If we have found the general points on the efficient frontier, (root{0.0111+1.35c^2}, 0.05+0.27c), why isn't the gradient just: gradient = (0.05+0.27c - 0.04) / (root{0.0111+1.35c^2} - 0) ?
The new efficient frontier (EF) is a line that runs TANGENTIAL to the old EF. Just by joining the risk-free asset with any point on the old EF doesn't necessarily get you a tangential line, John
Gradient question Yes, something about not knowing c. In truth, we can use the equation you, Jensen, put together, by finding the c which maximises it, since this will give the maximum return per unit risk, which is what we want. However, I think the solution would be faster to do in the exam, so worth learning it!
From the formula for a portfolio expected return, i.e. proportion invested in asset A multiplied by expected return on Asset A etc, using the form of effecient portfolios provided in the question as the proportions.