Guy proves that Buy The Dip strategy works (via backtests) any rebuttals?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
protagonist
Posts: 4952
Joined: Sun Dec 26, 2010 12:47 pm

Re: Guy proves that Buy The Dip strategy works (via backtests) any rebuttals?

Post by protagonist » Fri Sep 29, 2017 8:42 am

financeidiot wrote:
Thu Sep 28, 2017 9:56 am
You're probably going to get the same set of rebuttals to every response about a market timing strategy that appears to work:
1. Past performance is not indicative of future performance. Models adjusted to fit backtested data are not guaranteed to fit future data.
2. Historical stock market data are limited. The OP seems to have backtested up to 40 years where market data are available. However, 40 years in the history of global economics is limited.
3. Transaction costs and taxes lower performance. The original poster was using a tax-free account in Canada to execute trades, this leads to better performance than if the strategy had taken place in a U.S. taxable account, performance would have been lower (but still good).
4. Effective strategies become less effective over time as more people discover and adopt them.
5. Non-passive strategies are psychologically difficult to execute. The OP has above average fortitude to stick to buying the dips. Most people cannot stay dedicated to a similar strategy.
6. Non-passive strategies are technically difficult to execute. This strategy appears more straightforward, but most people cannot execute complicated trading strategies.

The few big winners of investing in the future will be active traders, the many big losers will be active traders, the happy middle will be Bogleheads. There's always going to be some strategy that is better than passive investing, but most people will not be able to identity it, execute it, and abandon it when the time is right.
Adding to all this.....we have not had a prolonged "slow bleed" for a long time, where a market in decline continues to decline.

When things rebound rapidly, of course the strategy works. But if they don't you can be in serious trouble. And you never know.

When the NASDAQ fell from around 5000 to around 3500 around 1999 I thought it was a "great buying opportunity" . By 2002 it was down around 1100. Accounting for inflation I don't know if the NASDAQ has rebounded in real terms to its 1999 high yet.

So the strategy comes with serious risks.

Just think "Japan". If you kept buying the Nikkei on dips you would be broke by now.

Nobody "proves" anything by using limited past data.
Last edited by protagonist on Fri Sep 29, 2017 8:47 am, edited 1 time in total.

User avatar
rocket354
Posts: 119
Joined: Mon Dec 14, 2015 12:31 pm

Re: Guy proves that Buy The Dip strategy works (via backtests) any rebuttals?

Post by rocket354 » Fri Sep 29, 2017 10:08 am

goingup wrote:
Fri Sep 29, 2017 8:28 am

:wink: Well really contributing every pay period to a 401K is called periodic investing which is often conflated with dollar cost averaging. It's definitely not lump sum investing.
Well, sorry to continue this sidebar, but I am curious how it's different from lump-sum (legitimately, not argumentatively). I'd be putting every dollar I can into the market at the moment I can. What happened two weeks ago, or will happen in two weeks doesn't seem relevant.

Or is it the 401k vs taxable distinction? A 401k investment isn't necessarily maxed out, it's (often) designed to be contributed over the year to always get the employer match, when someone could theoretically contribute more. Because of the intent and reasons, that would be considered periodic rather than DCA. However, the original comment was about DCAing into investment accounts, non-specified. If I contribute as much as I can at the moment I get it, that to me is lump sum, even if it is a small sum. The 401k portion may be periodic (or may not be, depending on my salary, and structure of the match), and its existence is being assumed, but the overall contribution still seems lump-sum to me.

livesoft
Posts: 57272
Joined: Thu Mar 01, 2007 8:00 pm

Re: Guy proves that Buy The Dip strategy works (via backtests) any rebuttals?

Post by livesoft » Fri Sep 29, 2017 10:35 am

I really want to encourage the OP to try this and tell us how it works out. No cheating though.
This signature message sponsored by sscritic: Learn to fish.

User avatar
triceratop
Moderator
Posts: 3784
Joined: Tue Aug 04, 2015 8:20 pm
Location: la la land

Re: Guy proves that Buy The Dip strategy works (via backtests) any rebuttals?

Post by triceratop » Fri Sep 29, 2017 10:42 am

MarginalCost wrote:
Thu Sep 28, 2017 8:36 pm
livesoft wrote:
Thu Sep 28, 2017 9:04 am
I am a proponent of good market timing. I buy when prices drop in a specified way and rebalance shortly thereafter*. There has been backtesting performed as shown in this thread where many similar backtests fail, but a few are actually quite good. The trick is to be emotionally ready to not do the "similar" ones and to do the ones that work. I think very few people can do that emotionally.

Also beware of outliers. That is, where one single trade can make or break the entire analysis. One needs to do some cross-validation.

In the end no one will believe you anyways, until you have a blog that has done it for at least 10 years and outperformed. Hence, my public blog with my trades published.

*Those are NOT the only times that I buy nor the only times that I rebalance.
Livesoft, where is this blog? I would love to see it
This is an old joke, or at least one I've seen several times (and never explained); livesoft treats bogleheads.org as his personal blog. :D
"To play the stock market is to play musical chairs under the chord progression of a bid-ask spread."

goingup
Posts: 2837
Joined: Tue Jan 26, 2010 1:02 pm

Re: Guy proves that Buy The Dip strategy works (via backtests) any rebuttals?

Post by goingup » Fri Sep 29, 2017 11:21 am

deleted

angelescrest
Posts: 889
Joined: Tue May 27, 2008 10:48 am
Location: The Third Coast

Re: Guy proves that Buy The Dip strategy works (via backtests) any rebuttals?

Post by angelescrest » Sat Sep 30, 2017 8:05 am

flyingaway wrote:
Thu Sep 28, 2017 10:34 am
I wish I could know what is coming is a dip or a pre-crash.
Me, too, but I am really smart and would buy it once I know the crash hits bottom. :P

User avatar
nedsaid
Posts: 8831
Joined: Fri Nov 23, 2012 12:33 pm

Re: Guy proves that Buy The Dip strategy works (via backtests) any rebuttals?

Post by nedsaid » Sat Sep 30, 2017 2:59 pm

I didn't want to be too snarky when commenting about future testing. We can always come up with an optimal strategy that would have worked perfectly in the past. The economy and the markets are dynamic and thus what worked in the past may not work in the future. There are limits to what back testing can accomplish.

That being said, buying on the dips is not a bad idea. It is always better to buy stocks at a cheaper price than a more expensive price. The problem is, you might wait a long time before the market drops enough for you to make a new investment. You might be better off just investing at regular intervals regardless of price. It is just that markets do what markets do and not necessarily what we want them to do. Even if the markets do what we want them to, they may not do what we want when we want. Another factor, is that even now, cash does not return very much and having money sit in a low return asset class for long periods of time would be a drag on your returns.

If you are a young investor, I would just dollar cost average my investments in over time. Most people with 401(k)'s are forced to do this as contributions come out of your paychecks. Don't sweat it.
A fool and his money are good for business.

hoops777
Posts: 1997
Joined: Sun Apr 10, 2011 12:23 pm

Re: Guy proves that Buy The Dip strategy works (via backtests) any rebuttals?

Post by hoops777 » Sat Sep 30, 2017 4:16 pm

snarlyjack wrote:
Thu Sep 28, 2017 7:34 pm
I dca into my accounts every payday.
I don't care what the markets are doing (up, down, sideways).
I' am a long term buy & hold investor. It's time in the
markets not timing the markets...that's the big secret.
You buy 100,000 of SP500 Monday.I buy 100,000 1 month from today after it tanked 20 pct.Who has more money 20 years from today?Just making a point that timing CAN make a difference.We all know the opposite effect as well so no lectures needed. :D
K.I.S.S........so easy to say so difficult to do.

lotusflower
Posts: 102
Joined: Thu Oct 24, 2013 12:32 am

Re: Guy proves that Buy The Dip strategy works (via backtests) any rebuttals?

Post by lotusflower » Sat Sep 30, 2017 6:48 pm

hoops777 wrote:
Sat Sep 30, 2017 4:16 pm
You buy 100,000 of SP500 Monday.I buy 100,000 1 month from today after it tanked 20 pct.Who has more money 20 years from today?Just making a point that timing CAN make a difference.We all know the opposite effect as well so no lectures needed. :D
Well if someone buys SP500 EVERY Monday, and someone else only buys after 20% dips keeping the rest in CDs, then after 20 years I'm pretty sure the former will have come out ahead.

hoops777
Posts: 1997
Joined: Sun Apr 10, 2011 12:23 pm

Re: Guy proves that Buy The Dip strategy works (via backtests) any rebuttals?

Post by hoops777 » Sat Sep 30, 2017 8:50 pm

lotusflower wrote:
Sat Sep 30, 2017 6:48 pm
hoops777 wrote:
Sat Sep 30, 2017 4:16 pm
You buy 100,000 of SP500 Monday.I buy 100,000 1 month from today after it tanked 20 pct.Who has more money 20 years from today?Just making a point that timing CAN make a difference.We all know the opposite effect as well so no lectures needed. :D
Well if someone buys SP500 EVERY Monday, and someone else only buys after 20% dips keeping the rest in CDs, then after 20 years I'm pretty sure the former will have come out ahead.
Your comment missed my point and is irrelevant to what I said.
K.I.S.S........so easy to say so difficult to do.

flyingaway
Posts: 1532
Joined: Fri Jan 17, 2014 10:19 am

Re: Guy proves that Buy The Dip strategy works (via backtests) any rebuttals?

Post by flyingaway » Sat Sep 30, 2017 11:16 pm

angelescrest wrote:
Sat Sep 30, 2017 8:05 am
flyingaway wrote:
Thu Sep 28, 2017 10:34 am
I wish I could know what is coming is a dip or a pre-crash.
Me, too, but I am really smart and would buy it once I know the crash hits bottom. :P
Please PM me if the real bottom is coming in the next crash. I just need one to retire.

investorpeter
Posts: 42
Joined: Sun Jul 31, 2016 5:46 pm

Re: Guy proves that Buy The Dip strategy works (via backtests) any rebuttals?

Post by investorpeter » Sun Oct 01, 2017 12:08 am

The problem as others have said is this is just a backtest. This is curve fitting using the entire data set. Every set of data will have an optimal curve that fits best on backtesting. So over this period of time, for this stock, for these indicators and variables, the optimal curve was found. What should have been done was to divide the data into a training set and a testing set. Just looking at the training set a strategy is devised. Then use that strategy on the testing set. Then do this over and over dividing up the training and test sets in different ways. Then collect realtime data for a period of time and see if it still works. If you find a strategy that works through all of this, then you can try it it with real money. But I doubt this strategy would have gotten this far. These are very common momentum indicators that I'm sure others have looked at and tried. And if it did work, it would only work temporarily until others figured it out and started using the same strategy, or worse, figured out a way to make a lot more money by taking the other side of all the people blindly using the same strategy.

technovelist
Posts: 2696
Joined: Wed Dec 30, 2009 9:02 pm
Contact:

Re: Guy proves that Buy The Dip strategy works (via backtests) any rebuttals?

Post by technovelist » Sun Oct 01, 2017 4:02 am

investorpeter wrote:
Sun Oct 01, 2017 12:08 am
The problem as others have said is this is just a backtest. This is curve fitting using the entire data set. Every set of data will have an optimal curve that fits best on backtesting. So over this period of time, for this stock, for these indicators and variables, the optimal curve was found. What should have been done was to divide the data into a training set and a testing set. Just looking at the training set a strategy is devised. Then use that strategy on the testing set. Then do this over and over dividing up the training and test sets in different ways. Then collect realtime data for a period of time and see if it still works. If you find a strategy that works through all of this, then you can try it it with real money. But I doubt this strategy would have gotten this far. These are very common momentum indicators that I'm sure others have looked at and tried. And if it did work, it would only work temporarily until others figured it out and started using the same strategy, or worse, figured out a way to make a lot more money by taking the other side of all the people blindly using the same strategy.
Exactly.

This isn't rocket surgery. :D
In theory, theory and practice are identical. In practice, they often differ.

User avatar
grayfox
Posts: 4617
Joined: Sat Sep 15, 2007 4:30 am
Location: Earth

Re: Guy proves that Buy The Dip strategy works (via backtests) any rebuttals?

Post by grayfox » Sun Oct 01, 2017 4:31 am

Vision wrote:
Thu Sep 28, 2017 8:56 am
https://www.reddit.com/r/investing/comm ... hilosophy/

I like the analysis this person did and how he came to conclusion that Buy The Dip strategy actually works.

Yes, it is market timing, but he also posted cold, hard facts of how it outperformed SPY.

What would be your rebuttals to this argument?
The scientific method cannot establish truths; it can only falsify hypotheses. Examples:

Hypthothesis 1: You can withdraw 10% from S&P500 portfolio for 30 years and never run out of money.
Easily proven false by looking at past returns, i.e. backtesting.

Hypthothesis 2: You can withdraw 1% from S&P500 portfolio for 30 years and never run out of money.
Can not be proven false by backtesting. But it can not be proven to be true.

https://www.youtube.com/watch?v=OL6-x0modwY
Gott mit uns.

Da5id
Posts: 1637
Joined: Fri Feb 26, 2016 8:20 am

Re: Guy proves that Buy The Dip strategy works (via backtests) any rebuttals?

Post by Da5id » Wed Oct 04, 2017 8:01 am

grayfox wrote:
Sun Oct 01, 2017 4:31 am
Hypthothesis 2: You can withdraw 1% from S&P500 portfolio for 30 years and never run out of money.
Can not be proven false by backtesting. But it can not be proven to be true.
No No No! You are a heretic. The 4% rule is holy writ, a law of nature that can't under any circumstances be questioned or challenged, the science is settled here. Burning at the stake is too good for you, you'll be sentenced to an eternity of listening to Dave Ramsey explaining how you'll get a 10% return going forward.

Anyway, yeah, good post.

User avatar
zaboomafoozarg
Posts: 1853
Joined: Sun Jun 12, 2011 12:34 pm

Re: Guy proves that Buy The Dip strategy works (via backtests) any rebuttals?

Post by zaboomafoozarg » Wed Oct 04, 2017 8:24 am

Da5id wrote:
Wed Oct 04, 2017 8:01 am
grayfox wrote:
Sun Oct 01, 2017 4:31 am
Hypthothesis 2: You can withdraw 1% from S&P500 portfolio for 30 years and never run out of money.
Can not be proven false by backtesting. But it can not be proven to be true.
No No No! You are a heretic. The 4% rule is holy writ, a law of nature that can't under any circumstances be questioned or challenged, the science is settled here. Burning at the stake is too good for you, you'll be sentenced to an eternity of listening to Dave Ramsey explaining how you'll get a 10% return going forward.
Nah, I think a lot of people realize now that 4% is too risky. 2.5% is the new 4%.

Post Reply