I suggest that the accuracy of target prices should be measured with some relevant metric. The metric for accuracy could be, for example, Mean absolute percentage error (MAPE).
\\begin{align}\ne_{avg}=\\frac{1}{n}\\left(\\sum_{t=1}^n \\frac{\\lvert s_t-p_t\\rvert}{s_t}\\right)\\cdot100\\%\n\\end{align}
e_{avg} = mean absolute percentage error (MAPE)
n = number of data points
s_t = actual signal value at time t = actual share price
p_t = predicted signal value at time t = target price
@Juha_Kinnunen and @Antti_viljakainen, what is your view on systematically measuring accuracy and making the metrics visible?
P.S. It’s super great to be part of this awesome community! Many thanks to the Inderes team and the excellent community.