Author Topic: Backtest of only buying after a 5-10-15-20% correction?  (Read 6696 times)

Gone Fishing

  • Magnum Stache
  • ******
  • Posts: 2786
  • So Close went fishing on April 1, 2016
    • Journal
Backtest of only buying after a 5-10-15-20% correction?
« on: October 16, 2015, 01:04:00 PM »
I was trying to find a backtest of stockpiling cash then only buying after a correction vs DCA, but can't seem to find anything.  My gut is that DCA would win, but I'm trying to find by how much.

Jeremy E.

  • Handlebar Stache
  • *****
  • Posts: 1946
  • Location: Lewiston, ID
Re: Backtest of only buying after a 5-10-15-20% correction?
« Reply #1 on: October 16, 2015, 01:19:16 PM »
I was trying to find a backtest of stockpiling cash then only buying after a correction vs DCA, but can't seem to find anything.  My gut is that DCA would win, but I'm trying to find by how much.
What's most important is that historically, frontloading is the winner

StressLess

  • 5 O'Clock Shadow
  • *
  • Posts: 57
Re: Backtest of only buying after a 5-10-15-20% correction?
« Reply #2 on: October 16, 2015, 01:21:10 PM »
would be great to see if you find it....

personally, I dont think that DCA would do better BUT in reality it is impossible to:

A) know when a correction is coming and wait to buy
B) know when the correction is over and then buy

If one could correctly time the market (and consistently too), I don't think there is any dispute that one can outperform DCA or longterm and low fee index investing. but that is a GIANT "IF"....

i think thats the kind of thinking that got me into trouble in the past with investing.  thinking 'maybe ill be lucky'  'if i just wait on the sidelines for a little longer'

never happens, you'll get in and out at the exact wrong times and the stress will drive you insane

sirdoug007

  • Pencil Stache
  • ****
  • Posts: 587
  • Age: 39
  • Location: Houston, TX
Re: Backtest of only buying after a 5-10-15-20% correction?
« Reply #3 on: October 16, 2015, 01:24:29 PM »
Have you seen this (and the linked spreadsheets at the bottom of the post)?  http://www.ritholtz.com/blog/2014/08/worlds-greatest-market-timer/

In the US DCA wins out over the super trader with perfect timing.

Not sure about specifically your XX% correction question but I would start with the Shiller S&P500 data and make a spreadsheet to backtest it yourself.

Gone Fishing

  • Magnum Stache
  • ******
  • Posts: 2786
  • So Close went fishing on April 1, 2016
    • Journal
Re: Backtest of only buying after a 5-10-15-20% correction?
« Reply #4 on: October 16, 2015, 01:25:07 PM »
I was trying to find a backtest of stockpiling cash then only buying after a correction vs DCA, but can't seem to find anything.  My gut is that DCA would win, but I'm trying to find by how much.
What's most important is that historically, frontloading is the winner

Sure, but I am referring to a typical working investor that would save and invest as their wages are earned, not someone with a lump of cash on hand.

Gone Fishing

  • Magnum Stache
  • ******
  • Posts: 2786
  • So Close went fishing on April 1, 2016
    • Journal
Re: Backtest of only buying after a 5-10-15-20% correction?
« Reply #5 on: October 16, 2015, 01:27:20 PM »
would be great to see if you find it....

personally, I dont think that DCA would do better BUT in reality it is impossible to:

A) know when a correction is coming and wait to buy
B) know when the correction is over and then buy

If one could correctly time the market (and consistently too), I don't think there is any dispute that one can outperform DCA or longterm and low fee index investing. but that is a GIANT "IF"....

i think thats the kind of thinking that got me into trouble in the past with investing.  thinking 'maybe ill be lucky'  'if i just wait on the sidelines for a little longer'

never happens, you'll get in and out at the exact wrong times and the stress will drive you insane

That's why I would like to see what would happen if the "trigger" was set at a fixed percentage drop in the market.  No thinking involved, just throwing the lump in when the market corrects by x%, every time, over an extended period of time.

Gone Fishing

  • Magnum Stache
  • ******
  • Posts: 2786
  • So Close went fishing on April 1, 2016
    • Journal
Re: Backtest of only buying after a 5-10-15-20% correction?
« Reply #6 on: October 16, 2015, 01:40:12 PM »
Have you seen this (and the linked spreadsheets at the bottom of the post)?  http://www.ritholtz.com/blog/2014/08/worlds-greatest-market-timer/

In the US DCA wins out over the super trader with perfect timing.

Not sure about specifically your XX% correction question but I would start with the Shiller S&P500 data and make a spreadsheet to backtest it yourself.

Thanks!  That's as close as I've seen to what I am looking for.  I wonder what the returns were on his cash while he was waiting to invest, and how many times he actually ended up investing.  Would lowering the buying threshold actually result in a better return?

Vilgan

  • Bristles
  • ***
  • Posts: 451
  • Location: Seattle, WA
Re: Backtest of only buying after a 5-10-15-20% correction?
« Reply #7 on: October 16, 2015, 03:03:41 PM »
Have you seen this (and the linked spreadsheets at the bottom of the post)?  http://www.ritholtz.com/blog/2014/08/worlds-greatest-market-timer/

In the US DCA wins out over the super trader with perfect timing.

Not sure about specifically your XX% correction question but I would start with the Shiller S&P500 data and make a spreadsheet to backtest it yourself.

That's an awesome article! going to bookmark it for future reference when discussing this with people, thanks!

Jeremy E.

  • Handlebar Stache
  • *****
  • Posts: 1946
  • Location: Lewiston, ID
Re: Backtest of only buying after a 5-10-15-20% correction?
« Reply #8 on: October 16, 2015, 03:22:21 PM »
Have you seen this (and the linked spreadsheets at the bottom of the post)?  http://www.ritholtz.com/blog/2014/08/worlds-greatest-market-timer/

In the US DCA wins out over the super trader with perfect timing.

Not sure about specifically your XX% correction question but I would start with the Shiller S&P500 data and make a spreadsheet to backtest it yourself.

Thanks!  That's as close as I've seen to what I am looking for.  I wonder what the returns were on his cash while he was waiting to invest, and how many times he actually ended up investing.  Would lowering the buying threshold actually result in a better return?
I think some people have this as a strategy to reduce their bond percentages while raising their stock percentages. I think go curry crackers plan is "If a 25% drop happens, then Iíll move 1/2 of our bond position into VTI. If the market drops another 25%, Iíll move the other 1/2" However I think it would make more sense for them to sell all their bonds today and put them into VTI, just as I think it makes more sense to buy as soon as funds are available. The market may never go lower than it is today.

steveo

  • Handlebar Stache
  • *****
  • Posts: 1943
Re: Backtest of only buying after a 5-10-15-20% correction?
« Reply #9 on: October 16, 2015, 04:29:01 PM »
The market may never go lower than it is today.

This is a big call to make. I'm reading some of Bernstein's books and he has stated how people do this and then the market stagnates for years.

Jeremy E.

  • Handlebar Stache
  • *****
  • Posts: 1946
  • Location: Lewiston, ID
Re: Backtest of only buying after a 5-10-15-20% correction?
« Reply #10 on: October 16, 2015, 04:41:56 PM »
The market may never go lower than it is today.

This is a big call to make. I'm reading some of Bernstein's books and he has stated how people do this and then the market stagnates for years.
Let me rephrase, the market valuation might never be less than what it is today. Of course it's possible that the market stagnates, drops, etc. But historically it's been best to put funds in as they are available. If the market is valued at 2000 today, maybe tomorrow it'll be 2020, and go up everyday until it's at 2400, then the next day it drops to 2100 and everyone goes crazy and then you buy, you're still buying higher than 2000. This is hypothetical. Historical returns are not a guarantee of future returns.

bryan

  • 5 O'Clock Shadow
  • *
  • Posts: 50
Re: Backtest of only buying after a 5-10-15-20% correction?
« Reply #11 on: October 16, 2015, 07:59:11 PM »
I was trying to find a backtest of stockpiling cash then only buying after a correction vs DCA, but can't seem to find anything.  My gut is that DCA would win, but I'm trying to find by how much.

DCA is the obvious winner for the simple case (white collar worker).

But more interesting, to me, is the capability to backtest asset allocation switching strategies. You can model something like your question e.g. 95% TSM DCA but if 5-10-15-20% correction, take your lquid allocation and go 100% TSM.

You could come up with some better ones though (e.g. golden butterfly v TSM v permanent portfolio).

MDM

  • Walrus Stache
  • *******
  • Posts: 9894
Re: Backtest of only buying after a 5-10-15-20% correction?
« Reply #12 on: October 17, 2015, 12:41:47 AM »
I was trying to find a backtest of stockpiling cash then only buying after a correction vs DCA, but can't seem to find anything.  My gut is that DCA would win, but I'm trying to find by how much.
Not sure if www.portfoliovisualizer.com will meet your need, but if you haven't checked it....


mrpercentage

  • Handlebar Stache
  • *****
  • Posts: 1236
  • Location: PHX, AZ
Re: Backtest of only buying after a 5-10-15-20% correction?
« Reply #14 on: October 17, 2015, 01:40:50 AM »
I was trying to find a backtest of stockpiling cash then only buying after a correction vs DCA, but can't seem to find anything.  My gut is that DCA would win, but I'm trying to find by how much.

I don't know of anything out there but I can tell you what I think. First, waiting for every correction as a general strategy is a bad idea. That doens't mean throwing significant portions into a correction is a bad. I think its a great idea but it shouldn't be the whole strategy.

I was recently playing with portfolio visualizer and discovered some interesting results. For example, if you drop $10,000 into PAMCX at year 2000 and dont add a thing (doing the same with the S&P500) PAMCX blows the 500 out of the water. If however, you start with $100 and dollar-cost-average $100 a month for 15 years then the 500 closes the gap significantly.

I have noticed a lot of dividend stocks seem to do much better dumping in a large sum at once. Timing is everything though. I would time it but I dont have the capital now.

As far as timing goes, I think I have some ability. Im 30th place right now on Barclays US Market Insight Challege. Not bad in my book. Im up 11%. I did it with Ford. If only I had a real million to play with I would have just made $110,000 in a month. Instead I just dumped $7,000 into a down payment for a car (a Ford go figure). By my  calculations it cost me about $130,000 in 30 years but hell-- YOLO. Atleast it wasnt everything. Some people have that on a credit card. I did use my credit card for the down payment for the points. I hate making credit card payments that huge. Hate it. I do love the car though.

There. You've seen my brain.

ShoulderThingThatGoesUp

  • Magnum Stache
  • ******
  • Posts: 3057
  • Location: Emmaus, PA
Re: Backtest of only buying after a 5-10-15-20% correction?
« Reply #15 on: October 17, 2015, 05:57:11 AM »
Have you seen this (and the linked spreadsheets at the bottom of the post)?  http://www.ritholtz.com/blog/2014/08/worlds-greatest-market-timer/

In the US DCA wins out over the super trader with perfect timing.

Not sure about specifically your XX% correction question but I would start with the Shiller S&P500 data and make a spreadsheet to backtest it yourself.

This is fantastic. I sent it to a brother-in-law who was annoyed at brother-in-law 2 because BIL2 convinced him to put his big cash pile into Vanguard just before the recent drop.

Proud Foot

  • Handlebar Stache
  • *****
  • Posts: 1080
Re: Backtest of only buying after a 5-10-15-20% correction?
« Reply #16 on: October 20, 2015, 11:41:44 AM »
but I would start with the Shiller S&P500 data and make a spreadsheet to backtest it yourself.

This right here.  It would probably take about 15 minutes to make a spreadsheet that would do this.  The question then is what time-frame are you looking at for your correction? Daily, weekly, monthly, some arbitrary number?

Gone Fishing

  • Magnum Stache
  • ******
  • Posts: 2786
  • So Close went fishing on April 1, 2016
    • Journal
Re: Backtest of only buying after a 5-10-15-20% correction?
« Reply #17 on: October 20, 2015, 12:35:27 PM »
but I would start with the Shiller S&P500 data and make a spreadsheet to backtest it yourself.

This right here.  It would probably take about 15 minutes to make a spreadsheet that would do this.  The question then is what time-frame are you looking at for your correction? Daily, weekly, monthly, some arbitrary number?

I'm not that smart!  But let's say daily.

Proud Foot

  • Handlebar Stache
  • *****
  • Posts: 1080
Re: Backtest of only buying after a 5-10-15-20% correction?
« Reply #18 on: October 21, 2015, 07:43:25 AM »
The problem with daily is that it doesn't happen that often unless you're looking at a narrow market.  I just pulled the history for VTSAX since November 13, 2000 and there were only 10 times it dropped over 5% in one day.  And those days were 9/29/08, 10/7/08, 10/9/08, 10/15/08, 10/22/08, 11/5/08, 11/12/08, 11/19/08, 11/20/08, and 12/1/08.  So if you're piling up your cash and dividends (assuming adding to cash on the 1st of each month) then you would only be making purchases 4 times during the ~15 year period.

Gone Fishing

  • Magnum Stache
  • ******
  • Posts: 2786
  • So Close went fishing on April 1, 2016
    • Journal
Re: Backtest of only buying after a 5-10-15-20% correction?
« Reply #19 on: October 21, 2015, 08:32:05 AM »
The problem with daily is that it doesn't happen that often unless you're looking at a narrow market.  I just pulled the history for VTSAX since November 13, 2000 and there were only 10 times it dropped over 5% in one day.  And those days were 9/29/08, 10/7/08, 10/9/08, 10/15/08, 10/22/08, 11/5/08, 11/12/08, 11/19/08, 11/20/08, and 12/1/08.  So if you're piling up your cash and dividends (assuming adding to cash on the 1st of each month) then you would only be making purchases 4 times during the ~15 year period.

Sorry, I misunderstood your question, let's say from the peak.  As someone else suggested, the 52 week high would probably work fine.

Proud Foot

  • Handlebar Stache
  • *****
  • Posts: 1080
Re: Backtest of only buying after a 5-10-15-20% correction?
« Reply #20 on: October 22, 2015, 08:22:30 AM »
So Close,

Here is the results I got.  I only tested VTSAX with the assumptions of buying after a 5%, 10%, and 15% drop from 52 week  high compared to immediate investment. My assumptions were you had $10,000 for to invest at day 1 (November 13, 2000) and had an additional $1,000 of cash available for investment on the first day (day markets were opened) of each month.

Investing after a 5% drop beat the DCA marginally over the 15 year test period. 
The final balances and XIRR were:
5%     $372,671.80   6.62$
10%   $347,166.59   6.04%
15%   $346,353.09   6.03%
None  $372,037.91    6.57%

Obviously these results will vary depending on the fund you are investing in.
« Last Edit: August 20, 2017, 09:56:26 AM by Proud Foot »

Gone Fishing

  • Magnum Stache
  • ******
  • Posts: 2786
  • So Close went fishing on April 1, 2016
    • Journal
Re: Backtest of only buying after a 5-10-15-20% correction?
« Reply #21 on: October 22, 2015, 09:20:59 AM »
So Close,

Here is the results I got.  I only tested VTSAX with the assumptions of buying after a 5%, 10%, and 15% drop from 52 week  high compared to immediate investment. My assumptions were you had $10,000 for to invest at day 1 (November 13, 2000) and had an additional $1,000 of cash available for investment on the first day (day markets were opened) of each month.

Investing after a 5% drop beat the DCA marginally over the 15 year test period. 
The final balances and XIRR were:
5%     $372,671.80   6.62$
10%   $347,166.59   6.04%
15%   $346,353.09   6.03%
None  $372,037.91    6.57%

Obviously these results will vary depending on the fund you are investing in. I have attached my spreadsheet so you can look at and play with different assumptions

https://docs.google.com/spreadsheets/d/1BXHOcN0bVm3G1WXAMNP_rzO3QsuhleG-0qnrzIBy7II/edit?usp=sharing

Awesome!  Thank you! 

ditheca

  • Stubble
  • **
  • Posts: 244
  • Age: 36
  • Location: ST GEORGE, UT
Re: Backtest of only buying after a 5-10-15-20% correction?
« Reply #22 on: October 23, 2015, 12:00:43 PM »
Anyone else amused by the comment prompt at ritholtz.com?

Quote
Please use the comments to demonstrate your own ignorance, unfamiliarity with empirical data and lack of respect for scientific knowledge. Be sure to create straw men and argue against things I have neither said nor implied. If you could repeat previously discredited memes or steer the conversation into irrelevant, off topic discussions, it would be appreciated. Lastly, kindly forgo all civility in your discourse . . . you are, after all, anonymous.