Based on the reported literature and commonly used metrics in the realm of solar forecasting, a new methodology is developed for estimating a metric called forecastability (F). It reveals the extent to which solar radiation time series can be forecasted and provides the crucial context for judging the inherent difficulty associated to a particular forecast situation. Unlike the score given by the standard smart Persistence model, the F metric which is bounded between 0% and 100% is easier to interpret hence making comparisons between forecasting studies more consistent. This approach uses the Monte Carlo method and estimates F from the standard error metric RMSE and the Persistence predictor. Based on the time series of solar radiation measured at 6 very different locations (with optimized clear sky model) from a meteorological point of view, it is shown that F varies between 25.5% and 68.2% and that it exists a link between forecastability and errors obtained by machine learning prediction methods. The proposed methodology is validated for 3 parameters that may affect the F estimation (time horizon, temporal granularity and solar radiation components) and for 50 time series relative to McClear web service and to the central archive of Baseline Surface Radiation Network. 82 time series formalism is the result of the work of the co-83 recipient of the Nobel Prize Clive W.J. Granger 15. Authors 84 define it as the variance of the optimal forecast divided by the 85 unconditional variance of the time series. This definition and 86 the resulting Q parameter (forecastability quotient) were ex-87 tensively studied in economics, gradually giving way to new 88 kind tools as sample and approximate entropy 16 , correlation 89 and mutual information metrics 17. 90 These two notions (P and F) are conceptually very close: if 91 the predictability (P) 18 studies how trajectories of the true sys-92 tem diverge 19 , the forecastability (F) describes how a model 93 trajectory diverges from a true system trajectory 14. A common 94 explanation is that a predictable process is able to be predicted 95 while a forecastable one is able of being forecasted. With this 96 last definition, the concept of modelling appears, thereby a 97 forecastable system is necessarily predictable but the opposite 98 is not true ii. The predictability term which is often used with 99 dynamic processes, is closely related to notions like causality 100 21 or chaos (i.e. failure of predictability 22), found for example, 101 in all weather series and where the typical predictable times 102 (or barriers) concerns the prediction horizons smaller than 1 103 or 5 days 23. In the context of the present study (nowcasting or 104 very short term), the chaotic aspect is not directly studied, so, 105 the term forecastability seems more suitable than predictabil-106 ity one even if, to the best of our knowledge, there isn't any 107 consensus on that. 108 Note that other concepts could have been detailed here (ob-109 servability and detection reliability 24) but they do not have 110 much sense in the study of solar radiation given the quantities 111 involved are directly measurable and the associated uncertain-112 ties are very low.