The two key factors are timing and excessive force. The easiest way to produce a pulled anchor is to (1) start forward too early. Simply hit it hard just before the anchor can possibly land and it will kick out like a wild mule. Another is to use (2) too much force in either direction, ironically more likely the backcast from what I have observed.
Why?
The anchor relies on a subtle descent from flight to kiss the water. If you start the forward power stroke too early, this gentle descent never occurs, or at least not fully.
Details:
The force of the forward stroke must not reach the far end of the line before this descent begins ot it will force a continuation of flight (no descent) and pull the anchor out from under us like a rug when the rest of the forward power stroke comes into play. In other words the descent is cancelled out.
It takes a fraction of a second for the brunt of the forward stroke to reach the other end of the line after the rod moves forth, and this pause is among the most important elements of timing in all of Spey casting.
Likewise, if too much force is used in the backcast the same occurs because the bottom half of the d-loop will just keep flying without descent to the water, even if the timing is good - you wait for something that never occurs.
Case in point - snap-t / circle spey...
This setup motion is done by 'pulling' the line into a straight line movement toward the upriver direction, then 'pushing' the near end of the working length of line in the opposite direction. This opposing tension is so amplified that the line snaps tight above the water in the trademark, "chicks dig it" fashion (re: Dec Hogan nomenclature). This is in fact the same as an anchor kicking out in principle, except that a snap-t 'pushes underneath the pull' where a pulled anchor involves opposing tension above.
In fact I've observed that many folks just learning the snap-t for the first time mistakingly oppose to the top, which is natural because that's how we cast. What results is a forcefully pulled anchor!
Additional notes:
1) There are several other ways to cause it, for instance (3) an excessive lift at the end of the d-loop pulling the d-loop out of shape, making it unlikely a good anchor will result (unless the bottom hand trunks, common with high arm raisers which causes excess anchor). Any motion that disrupts or deforms the gentle descent of the bottom half of the d-loop will likely have a poor result when the forward stroke's power reaches it.
2) The anchor is such a critical element because it complements the weight and momentum of the line to support the 180 degree reversal of direction. In Spey casting, things take time to happen. Although measured in fractions of a second, the anchor affords momentary stability during the interval of time it takes for the d-loop to transform itself into the actual cast.
3) When considering too much anchor as well as too little - a shorter denser line has a wider margin of error in stroke timing than a longer line matched to the same rod. However it's easier to pull the anchor with the shorter line because of the application of power influences the other end of the line quickly. When casting long belly lines it's important to stay in sync with the anchor by starting forward early enough. Fortunately the added mass and resistance present with extended belly lines (when in motion) allows this and when the timing is right it all works out. Interestingly it's quite easy to pull the anchor on an extended belly line as well using the same principles, e.g.: snap-t.
From one instructor to another - good luck with your test.
