Fund managers, page-205

Hi @lost and@SP007

The reason our models are out is due to timing and the effect of compounding, I provided an explanation to Chella as he requested it on the MFG board under 'The horse has bolted'.

I wanted to share the response with you guys here as well, if you have any other ideas of why the model may be inaccurate please share.

I gave an two examples on a 10b fund (to make maths easy)
One of what the model would predict in terms of flows if there had been inflows and a positive market movement over the month.
Over the month the fund performs +2.00%
Suppose you put 100m in the fund after a week, after it had gone up 0.5%
The 100m is worth 101.5m at the end of the month (Increase of 1.5%)
If we didn't know the amount of inflows/outflows (Like the case we have with magellan)
We would calculate 10b x 1.02 and arrive at 10.2 billion as our estimate for the month
At the end of the month magellan reports FUM of 10.305 billion
We then assume there has been 101.5 million of inflows, however there was only 100 million, the reason is due to timing and performance over the period which distorts the numbers. This is relatively small but adds up on thousands of transactions over the month.


Another example but opposite (Outflows and negative performance)
Again assume a 10b fund Over the month the fund performs -3.00%
Suppose you redeem 100m in the fund after a week, after it had gone down 1%
The fund is expected to report a change from 10b to 9.7b,
The actual FUM reported would be is $9.604b instead of 9.7b, we then assume there has been 96m of outflows

What does this all mean?
The model over-estimates inflows and under-estimates outflows,
This is consistent with back testing
March quarter estimate 1826m of inflows, actual was 1120m of inflows (Over estimates inflows)
Sept quarter estimate 276m of inflows, actual was 1,527m of outflows (Under-estimate outflows)



The model predicts 1.434 b of outflows over OCT and NOV, This is the first time the model has predicted outflows and because we know the model under-estimates outflows i'd assume between 2 and 3 billion of outflows. In early Jan we'll have another data point for Dec FUM and all flows will be revealed.

I think the numbers will be quite ugly, it will be interesting to see if Magellan adds any commentary around further 'rebalancing' from institutions. And if the magellan posters come to the realisation of the change in direction of their business and stock.

Cheers,
Plague.