Equity Coverage

One simple question :

How many stocks can an analyst or a PM realistically cover before the quality of coverage for each one starts to decline ?

Say, cross-sectorial but one geography (Europe, Asia or North America)