WEBVTT

1
00:00:00.309 --> 00:00:02.909
Will $1,000,000 last
30 years of retirement?

2
00:00:04.608 --> 00:00:07.388
Start with a plan holding
a balance of $1,000,000.

3
00:00:08.578 --> 00:00:10.938
And 30 years of retirement
for it to cover.

4
00:00:11.639 --> 00:00:14.399
Monte Carlo runs one
plan through thousands of

5
00:00:14.399 --> 00:00:15.839
possible market futures.

6
00:00:16.699 --> 00:00:18.754
It reports how often
the plan survives

7
00:00:18.754 --> 00:00:20.539
and the range of
ending balances.

8
00:00:21.459 --> 00:00:23.816
As an example,
we will describe one plan,

9
00:00:23.816 --> 00:00:26.059
run it, and change
a single assumption.

10
00:00:27.670 --> 00:00:30.777
Next, choose how spending
responds to the portfolio,

11
00:00:30.777 --> 00:00:33.167
because a fixed rule
ignores the balance

12
00:00:33.167 --> 00:00:36.154
while guardrails and percentage rules cut spending

13
00:00:36.154 --> 00:00:37.170
after a bad year.

14
00:00:37.930 --> 00:00:39.846
Fixed dollar keeps
this test simple,

15
00:00:39.846 --> 00:00:41.230
so only the market varies.

16
00:00:42.099 --> 00:00:45.779
$40,000 a year, held
steady in real terms.

17
00:00:46.559 --> 00:00:49.356
Withdrawing 4% of the
starting balance is the rule

18
00:00:49.356 --> 00:00:50.699
of thumb this run tests.

19
00:00:52.395 --> 00:00:54.480
The tax-aware engine,
a Pro feature,

20
00:00:54.480 --> 00:00:57.377
routes every path through
your saved household for

21
00:00:57.377 --> 00:00:58.535
an after-tax result.

22
00:00:59.455 --> 00:01:00.842
This run stays pre-tax,

23
00:01:00.842 --> 00:01:03.435
so every figure that
follows is before tax.

24
00:01:05.609 --> 00:01:07.946
Open Advanced settings
to see the market model

25
00:01:07.946 --> 00:01:08.809
this run assumes.

26
00:01:09.429 --> 00:01:11.529
Returns are drawn from
a parametric model.

27
00:01:11.869 --> 00:01:14.169
Its annual average
return is 10%.

28
00:01:14.169 --> 00:01:16.609
Its annual
volatility is 15%.

29
00:01:16.609 --> 00:01:18.129
Both are before inflation.

30
00:01:19.089 --> 00:01:20.664
Using historical
return periods

31
00:01:20.664 --> 00:01:23.052
and a model with heavier
tails are alternatives

32
00:01:23.052 --> 00:01:25.339
when a smooth curve
fails to reflect crashes.

33
00:01:26.149 --> 00:01:28.477
Path count and inflation
sit here as well,

34
00:01:28.477 --> 00:01:30.917
and the recorded random
seed is what lets an

35
00:01:30.917 --> 00:01:32.469
identical run be reproduced.

36
00:01:33.813 --> 00:01:36.922
Measure how often this
plan survives the full

37
00:01:36.922 --> 00:01:40.446
horizon by running 10,000
possible return sequences

38
00:01:40.446 --> 00:01:41.413
over 30 years.

39
00:01:42.183 --> 00:01:44.425
Start with the
definition of success:

40
00:01:44.425 --> 00:01:47.577
a simulation succeeds when
money remains at year 30,

41
00:01:47.577 --> 00:01:49.335
while the failure
curve shows

42
00:01:49.335 --> 00:01:52.183
when unsuccessful simulations run out of money.

43
00:01:52.883 --> 00:01:55.711
The percentile band
represents many simulations

44
00:01:55.711 --> 00:01:58.658
rather than one by marking
the 10th-percentile or

45
00:01:58.658 --> 00:02:00.403
median balance
at each month.

46
00:02:01.471 --> 00:02:03.737
90% of those futures
still have money

47
00:02:03.737 --> 00:02:04.411
at the end.

48
00:02:05.191 --> 00:02:06.231
One in 10 does not.

49
00:02:08.189 --> 00:02:09.729
The headline
hides the range.

50
00:02:10.589 --> 00:02:12.949
The middle path
ends near $4.9M.

51
00:02:13.829 --> 00:02:16.829
Measured in the dollars
of year 30, not today's.

52
00:02:17.469 --> 00:02:20.138
The lowest 10% of
simulations end with almost

53
00:02:20.138 --> 00:02:22.689
nothing, even under
the same plan and rule.

54
00:02:23.871 --> 00:02:25.111
Failure has a shape.

55
00:02:26.091 --> 00:02:28.851
98% of futures are still
solvent at year 20.

56
00:02:29.631 --> 00:02:32.271
94% are still
solvent at year 25.

57
00:02:33.151 --> 00:02:35.501
The final decade contains
most of the failures.

58
00:02:36.696 --> 00:02:38.116
The withdrawal itself grows.

59
00:02:38.936 --> 00:02:40.734
Held steady in real terms,

60
00:02:40.734 --> 00:02:43.776
that $40,000 is about
$91,000 in the dollars

61
00:02:43.776 --> 00:02:44.536
of year 30.

62
00:02:45.266 --> 00:02:46.866
Save any run you
want to keep.

63
00:02:47.526 --> 00:02:48.514
It stores the plan,

64
00:02:48.514 --> 00:02:50.230
the market model,
and the result,

65
00:02:50.230 --> 00:02:52.726
so it can be reopened
later or compared against.

66
00:02:53.528 --> 00:02:56.389
Increase annual
spending by $10,000,

67
00:02:56.389 --> 00:02:59.568
producing a new total
of $50,000 a year.

68
00:03:00.428 --> 00:03:01.829
Keep the balance, horizon,

69
00:03:01.829 --> 00:03:03.068
and market model fixed.

70
00:03:03.788 --> 00:03:06.798
That raises the starting
rate from 4% to 5%;

71
00:03:06.798 --> 00:03:08.508
run the simulation again.

72
00:03:09.563 --> 00:03:11.411
$10,000 more per year,

73
00:03:11.411 --> 00:03:15.023
and the success rate
falls from 90% to 78%.

74
00:03:15.383 --> 00:03:17.249
The saved run
stays as it was,

75
00:03:17.249 --> 00:03:19.363
so you can update
it or keep both.

76
00:03:20.149 --> 00:03:23.589
All 10,000 simulations
share one parametric model.

77
00:03:24.409 --> 00:03:26.849
Its annual average
return is 10%.

78
00:03:26.849 --> 00:03:29.649
Its annual
volatility is 15%.

79
00:03:29.649 --> 00:03:31.289
Both are before inflation.

80
00:03:32.142 --> 00:03:33.822
One more thing
the average hides.

81
00:03:34.682 --> 00:03:36.122
Take the poor future from

82
00:03:36.122 --> 00:03:38.946
before and replay its own
returns in the opposite

83
00:03:38.946 --> 00:03:40.789
order with the
same withdrawals;

84
00:03:40.789 --> 00:03:43.382
it ends near $3.2M
instead of almost nothing.

85
00:03:44.002 --> 00:03:46.458
Losses early, while
money is being withdrawn,

86
00:03:46.458 --> 00:03:47.222
do the damage.

87
00:03:48.062 --> 00:03:49.712
Planners call that
sequence risk.

88
00:03:50.254 --> 00:03:52.415
Real markets can have
fatter tails than

89
00:03:52.415 --> 00:03:53.634
this parametric model.

90
00:03:54.434 --> 00:03:56.914
This run also omits
taxes and fees.

91
00:03:57.734 --> 00:04:00.317
The saved assumptions make
the result reproducible

92
00:04:00.317 --> 00:04:01.454
and open to challenge.

93
00:04:02.345 --> 00:04:04.645
Describe the plan and
choose a market model.

94
00:04:05.485 --> 00:04:07.475
Read success rate
beside outcome spread

95
00:04:07.475 --> 00:04:08.445
and failure timing.

96
00:04:09.485 --> 00:04:11.872
Save the run, change
one material assumption,

97
00:04:11.872 --> 00:04:13.145
and compare the results.

98
00:04:14.708 --> 00:04:16.948
Choose Take a tour for
a guided walkthrough.

99
00:04:17.508 --> 00:04:19.058
You can dismiss
it at any time.

100
00:04:22.019 --> 00:04:23.691
Open Docs for every setting,

101
00:04:23.691 --> 00:04:26.319
withdrawal rule,
and supported return model.

102
00:04:28.000 --> 00:04:29.297
Before you trust a plan,

103
00:04:29.297 --> 00:04:31.080
ask how many
futures it survives.

104
00:04:32.080 --> 00:04:33.700
Then go find the
input that breaks it.