Price prediction using historical trajectories

Was curious what it would look like if I simply scanned through all of history, look for price movements within the same 60-day window that are similar to current movements, and then see what happened to them 30 days later. The results are above.

Each line is a 90-day price trajectory. I used the first 60 days to fit the pattern to current price movements, and selected the top 14 where the patterns matched the closest. The decimal in the parenthesis is the range of the movement, denominated by the starting price (on day 0). The patterns are filtered such that only ranges of >15% are selected.

Some current caveats: is not a quant study - nothing very rigorous about this nor will I place any bets on it. Another caveat: I simply normalized the frame by min and max values - and simply filtered them to be > 15%, judge for yourselves how representative each move is with respect to the current one (35%).


Investing in stocks – better than holding cash?

While it is true that investing in equities yield better returns than holding cash, it does not mean that you will always do better investing than doing nothing.

When factoring in the realities of how people invest (not speaking about biases /flaws but simply how life works), the returns of stocks do not translate as well to investors' portfolios.

The main reason is that people tend to quote stock market returns (e.g. 150% gain from 10 years ago), but rarely do investors dump in their full investable amount in one shot. Rather, they slowly accrue investable cash over time, and they can only decide at that time how they want to deploy that cash.

Below I've simulated 3 portfolios: (1) a buy-on-the-dip strategy, (2) using dollar-cost averaging, and (3) simply not investing the cash.

Each start out with 10k investable cash, and accrue $50 of investable cash every trading day (≈ setting aside 1k each month to invest). The investable instrument is the S&P index.

For simplicity, dividends are not included in the returns, and this is offset by the cash portfolio not being invested in bonds. One of the reasons for simplicity is that this is not meant to be a research piece on how we should invest, but rather just to give a sense of what investing looks like in reality, with an appreciation of its worst-of-times where being fully invested over 15 years can approach the gains obtained by simply doing nothing (can you stomach that?).


Summary of findings:

  1. For the past 30 years, the difference between being invested vs doing nothing is only evident in the 12-year super bull run of 2009-2020
  2. Before that (from 1990 to 2008) any gains were largely wiped out by the downturns of the 2000 and 2008 (depending on when you started)
  3. DCA introduces more risk because it stays fully invested throughout, but yields better total returns; the Sharpe however is slightly lower than the buy-the-dip strategy (but the different is not big and retail investors like us rarely care for such a small difference in Sharpe; total returns to us is more of a priority)
  4. Comparing dollar-cost-averaging and buy-the-dip: in very bad downturns (2008), the decades of gains using the DCA can be easily wiped out, and strategy total returns can be the same as BTD
  5. Depending on when you started, the drawdowns have different impact (how much strategies 1 and 2 loses compared to the all-cash portfolio)

What this means:

  • Widen your range of expectations for future market returns (next 30 years), as we don't know if we can get the same 12-year bull run we had. Be prepared that investment returns may not be as good for the next 20 years or so.
  • There is no universal strategy that is better - see below charts to get a sense of the variabiliy of 'which strategy is better' - see how starting at different times can change your conclusions. That's just one variable; others are: initial capital, how to BTD, and most importantly what is the magnitude of the next bull run.
  • Stay invested anyway - the degree to that (full DCA or BTD) is up to you -  your psychological comfort and anticipated circumstances



  • Start with $10,000 cash, immediately invest to the target equity ratio (e.g. 70% equity)
  • Daily inflow of $50; if DCA, invest immediately, if not, save it for the next dip
  • If actual equity ratio is higher than target equity ratio, wait for next market recovery (makes new all-time-highs) to unwind (sell) equity positions at a rate of 0.1% of entire equity portfolio every day, and stop unwinding if market dips down again (back to buying mode)
  • Charts start at 1990, 1995, 2000, and 2005


Grey line: dollar-cost-averaging strategy
Black line: total portfolio value of equities (blue) and cash (dark green)
Lime green line: cash-only portfolio

From 1990

Notice how all 3 strategies go back to square 1 in 2008

From 1995

Now, both equity strategies fare worse than cash in 2008 just because we started 5 years late

From 2000

A worse picture

From 2005

Needles to show starting at 2010 - would have been spectacular

Refining Fire

This market downturn has taught me some things.

First, don't do bargain buys in any part of the bull cycle. I had to cut a few of these positions because the downturn was like a refining fire - only the good ones recover - the bad ones keep going down. Bargain buying in a bull cycle is tempting because everything else looks expensive, and there's thing hunt for yield / returns. But it's only a matter of time before the fire burns away the dross, leaving only the pure elements. It's more important to buy strong companies (even if they're expensive) rather than bargain buy weak companies (hoping they will yield a higher % return).

Second, percentage returns for the rallies are misleading - e.g. "oil is up 50% from its lows". It's not uncommon to see severely beaten down stuff rally in double digits, leading you to think you should have bought beaten down stuff. But we often forget we can hardly ever pick the bottom, and so these returns are only relevant to the minority who did catch the bottom. For the rest of us, we are likely to have endured some double digit drawdown, before seeing some double digit rally (from the bottom), leaving our position pretty much near breakevent.

Third, think carefully and clearly on the next steps. This is a WIP as I consider whether I should bet big on a down move, or start putting money back into equities. We have to give ourselves the leeway to be wrong about things. Some decisions are psychologically more difficult than others (e.g. make a bear call while it's rallying, and missing a once-in-a-decade opportunity to buy), but I have to weigh things out as I best can, and make the mental decision and not waver. Rigour is not always rewarded - accept that. Be content that you've made your best decision, regardless of outcome.

Covid19 Marketview

I'm taking the opportunity to think about the future and while I may not be the most qualified / knowledgeable, I think it's good practice to think about such things and pen them down anyway.

TLDR: I think the worst is still ahead of us, and currently the market is not pricing in the full potential effect of this crisis on the economy.


22 Mar 2020


  • Buy US equities later, prices expected to drop further
  • Buy those with a previous strong run, dominance in space, established, but haven taken a good hit
  • EM like India = Hold, China = Hold
  • Oil exposure ok to hold


Will US lose its dominance?

Considering the possibilities of a world power shift.

China is in a good position because they have had the backdrop of having that economic clout, and now they are recovering faster than US. But the world system is still quite entrenched, so I think it will take more than just this slowdown and who is faster to the recovery race to overthrow US. It will take a really crippling thing to happen to the US which renders their systems unusable and forcing the incumbent ‘clients’ to switch. Similar to how UK lost its dominance in WW2, US fortuitously gained. It was nothing UK could do. But that alone didn’t do it – it required a surge of innovation in the US to propel it forward.

So maybe this crisis is not enough. But it could be the start of something.

Maybe China will gain grounds it would never have the chance to gain until now. Not sure where though. It could further consolidate its power in being the engines of the world economy by looking to replace other links in supply chains that are now closed (because they were in Europe). Availability in a desperate time is valuable. It’s just too difficult to predict how, where, and to what magnitude this will happen.

Back to the question, I don’t think it will lose its dominance in the tech space nor will its stock market be left in the doldrums as the rest of the world accelerates. I do believe that the main indices will recover and make new ATHs in a matter of 5 years. That’s because this virus changes economic demand, and that is a temporary thing. It is not paradigm-altering. Consumer demand remains the same – people will still buy the same things after the crisis and use the same services.

I don’t think it will lose its dominance, though I do think that China will close the gap.


What are the prospects for emerging countries?

A year or a few years back, it was still a good thing to have because everything else in developed markets were expensive – i.e. both stocks and bonds, and we were hungry for yield. We have been since 2016 actually, leading to things like VIX apocalypse.

But now why would we bother with emerging economies when the developed ones are on discount? Looking at India vs US, both are down about 30% and so the relative value goes to US in terms of risk-reward. I doubt India can recover faster than the US.

China looks good given the above reasons, but it has only declined about 12% throughout this crisis, so in comparison to US it’s less of a priority.

I’m just going to keep IN and CN on a hold for now.


What’s the plan for loading up?

I expect this downturn to go further because the curve is just starting for US and probably in the middle or about 35% for Europe. Meaning there’s still some way to go which will really put a dampener on consumer demand and prices should reflect that accordingly.

It seems so far that the way the infection spreads is quite uniform in terms of trend in countries, so it’s not unreasonable to expect the same waves to occur in Europe then US.

Depending on how forward-looking the market is, these could have ALREADY been priced in at this -30% level. But only time will tell as the waves materialize whether the above is true i.e. if it’s not priced in then we’ll see further declines as the infections materialize. If it is already priced in, I expect to see some form of a bottom being carved out. Seeing a bounce in the midst of the wave could be something else (random optimism or some behaviour I really can’t explain) because I’ve yet to see a market so efficiently forward-looking.

Again, many market moves cannot be fully explained – I’ve been stumped quite a few times post-move even with the benefit of hindsight and a plethora of news attempting to explain things i.e. the explanations don’t make sense but more commonly no explanation is given “stocks move higher despite XXX” kind of rubbish headlines.


What to buy in US?

Major indices and tech companies. Generally, companies that have been solidly established and have further room to go but have been knocked back by this. Dominating tech companies are good buys to me, at these levels, as these will not be going away anytime soon. They also have the ability to tide through the storm.

Avoid stocks which are discount on discounts – i.e. prices have been falling anyway. A simple filter is to look for 10-year annualized total return and compare that against 1-mth or 2-mth change in price. With filter of some market cap.


What about oil?

I was a bit too hasty in getting into oil when it collapsed. It took one more hit from 30 to low 20s, which is my lower end of my range i.e. I really did not think it would get this bad.  I had underestimated the contango also – meaning I cannot roll over my positions without significant cost.

Though I think oil will rise back to more normal levels of $30, there is no way to play this as it is already priced that way for the Dec contracts.

Date USO WTI chg chg beta
Jun-14 38 106
Feb-16 8 36 -79% -66% 1.20
Sep-18 16 75 100% 108% 0.92
Mar-20 5 23 -69% -69% 0.99


USO looks decent though in riding spot oil. The contango is built in but at least that’s managed for me.

Funny thing about contango is that market is usually pretty correct – after sharp declines, further months are priced to a premium, and at peaks, they are priced to discount.

This is a reliable signal for the F1-F6 premium too.

What to buy in this recession

Just a quick thought - the last recession saw many financial institutions pulled under. But they were too important to the system, and so many never actually went away. Similarly in this crisis, perhaps we should be looking at the hardest hit sectors and ask: are their share prices reflecting the long term prospects accurately? I will be looking out for 'systemically important' hospitality / travel related companies that have been dragged down too much. It's only a matter of time before consumer demand, air travel, and mass events return. Like banks, they can never go away - they are too important, at least for this era.

Similarly, it would be interesting to see a repeat of 2008 in another sector - i.e. a major hospitality company fails and triggers bankruptcies and selloffs in other related companies and then spread to other less related companies.