I have read that the times of the trading day that have the best liquidity are at the open and close but these also have the most volatility. I have also heard the dictum “amateurs trade at the open and pros at close”. Is the best time to trade from 10:15 to 11:15 EST? I trade weekly on Monday. I had done some sensitivity analysis on my ports for open, avg and close prices and they all perform best with open. I am currently using the 10:15 rule but would like to hear if others have another time of day that works better for them. My universe is (NOOTC) and MktCap>30 and AvgDailyTot(200)>150000 AND Price>2 AND country(“CHN,HKG”)=FALSE AND Sector!=FINANCIAL
Any comments?
I use limit orders set at the current price and do not use AON. My intent is to insure I trade all the stocks no matter the current price or if in smaller lots.
David -
country(“CHN,HKG”)=FALSE AND Sector!=FINANCIAL
will get you good results in the past. Not sure about the future.
Steve
In the case of these two groups I am more concerned about their financial reporting opacity than I am about absolute returns for the portfolio. Their numbers scare me. Do you have a preferred time of day for trading to give you good liquidity and quick execution?
David,
I trade between 9:30 and 10:00 Monday morning almost exclusively (I don’t like to spend more time than that; I have a life outside of the market). I do have a few daily Ports so I obviously will trade them during the week. I set up all my trades before the market opens and only place market orders. Since I monitor the bid/ask delta I know within a fraction of a % at what price I will get filled. That way I will always get a fill and I won’t end up chasing the price with limit orders. In order to place my buy orders I need to try to place the sell orders first, but some of the sells will need to wait a little on the price movement and the bid/ask spread.
For buys; I trade a gap up as soon as the bid/ask difference is small, and the gap is less than 10% for micro & small caps and 5% for mid & large caps. I have tested for that and, on average, if the gap is larger than that the price will trend down during the day, so I will check it again the next day. If a stock is trending down or gaps down early I wait until a small uptick (about 0.1%) and then I place the trade. If the stock is moving sideways I place the trade as soon as the bid/ask difference is small. If by 10:00 the bid/ask delta has not narrowed sufficiently for the $ amount I want to trade, I will reject the stock and get a different one. I have found that if a stock doesn’t have sufficient volume to buy within 30 minutes, it probably won’t have the volume necessary to sell when I really need to (for example, when a stop loss is triggered). Although I use reasonable liquidity rules for the market cap range the Port is set up to trade, occasionally on the day of the trade then volume is not there. By checking the chart I can see that the stock had very high volume at some time period in the recent past but has insufficient volume now.
For sells; I trade a gap down as soon as the bid/ask delta is small, or I trade it quickly if the gap is larger than 10% for micro & small caps and 5% for mid & large caps. Unlike the gap up for a buy, a large gap down, on average, will continue down during the day. These gaps down will frequently recover some in a few days or weeks, but they have been sold in the Port so I need to sell them also and not guess on a recovery. Besides, I will need the cash from the sells to buy the recommendations. I have a daily Port that trades only gap down stocks and has been developed to buy and sell them at optimum times, so I don’t need to guess about one that gaps down in my regular ports. If the stock is trending up I wait until the stock has a small pull back of about 0.1% and then place the trade. If the price is moving sideways I trade as soon as the bid/ask delta is small.
By following the above guidelines I have been able to outperform my micro cap Ports and have a near zero slippage on the larger Ports. So far I am currently out performing all of my R2G Ports since they were launched. That is because they trade at the average of high & low, and use variable slippage.
Denny 
Denny,
thank you for the detailed response, very interesting. It sounds like you look closely at the bid/ask spreads and the direction the stock is taking and then just pull the trigger with a market order. I might want to try this.
Question: if you are trading mostly for just 30 minutes on a Monday and require yourself to trade everything per the port recommendations, how do you handle when things are not aligning for a buy/sell in that 30 minute timeframe? it sounds like you trade when the circumstances are right but that could fall outside of the 30 minute window on a Monday. How do you handle that?
I have been doing pretty much the opposite of you by chasing orders throughout a two hour period using limits. Slippage has been a problem with this approach. I do try to stay away from low liquidity stocks that have a broad bid/ask spread per my Universe.
I was not sure if time of day really mattered. Maybe not.
I appreciate the perspective.
David,
I seldom have to wait more than 30 minutes. All of the mid and large caps will have a narrow bid/ask delta within 5 or 10 minutes so I trade those first, and the Micro caps will usually follow within another 10 minutes. If not, like I said above, I reject that stock and get a new recommendation. But again, that seldom happens. That has so far always resulted in a stock that I could trade right away.
I think time of day DOES mater. You can prove it to yourself. Just run your Sim using next open, average of High & Low, and next close. You will find for almost all Sims that the next open sim out performs the avg of hi & low, and that out performs next close. That is why I trade as soon as the stocks meet my guidelines after the market opens.
Denny
Got it, makes sense.
I actually have run my sims earlier as experiments on time of day effect and it agrees with your comments on the morning being best. This seemed to be in conflict with what I had been reading on the web so was not sure. It looks like your experiences confirmed my tests.
[quoteSince I monitor the bid/ask delta I know within a fraction of a % at what price I will get filled.
…
Denny
[/quote]
Hi Denny thanks for sharing. Well thought out as usual. Which site do you get the “bid/ask delta” from?
Denny,
If you wouldn’t mind, I would be interested to hear what % of the stock price you consider to be a “small” bid/ask difference.
Thank you
This is what I look for in the bid/ask spread: For large caps, < 0.1% (usually 1 to 3 cents), mid caps < 0.2%, small caps < 0.3% and micro caps < 0.5%. I will be able to trade that within the first 30 minutes for all but a few stocks a month, and those will be microcaps (I have three 5 stock microcap Ports I trade).
Denny
Hi Denny:
I can see how you can know within a fraction of a % for higher volume mid and large caps stocks, but for lower volume stocks what steps do you take to avoid moving the price with market orders?
I’m guessing you might be limiting the market order to the size of current ask or bid size. But that can result in some rather small positions for lower volume stocks (often as low as 100 to 200 shares). Or are you using level 2 quots to gage the shares available within a slippage tolerance. If so, which trading platform are you using to get enough simulateous level 2 windows for all the stocks on your buy and sell lists?
I’m also curious about your experience with how “dark” pools affect your trading strategy. My current entry and exit strategy uses limit orders when a stock has less bid/offer sizes than I want. This is frequently the case for my small and micro cap portfolios. From time to time I can get a partial, even complete, fill by submitting a limit order at the last trade price if this is inbetween the current bid ask prices. At times I also get fills by putting in limit orders that split the difference between the bid and ask even when the last trade was done at the visible bid or ask. I’m curious of your perspective on what your market order approach gives up on finding “dark” pools.
Brian
Denny:
This sounds good, but it may require more screen real estate than I currently have. To watch for these 0.1% upticks on 10, 15 or 20 stocks, I’m assuming you have several monitors filled with charts for each stock on your buy and sell lists.
Or is there a better way? Are you using a trading platform that has pattern recognition to give you an alert if a down trending stock on you buy list has an uptick? And vice versa for stocks on your sell list.
Brian
Brian,
I trade on Fidelity’s Active Trader Pro platform. I use a single 24” monitor. On it I setup a table of all my trade recommendations. The table has a continuously running real time update of last, bid, & ask prices. On the same screen I have a 10 day chart, a table of level 2 quotes and a small window of news. If I double click on any stock in the trade table it’s displayed in the chart, the level 2 quote table, and the news. There also will be a popup during the day of any new news for any stock that I am holding.
On the trade table I set up the order of the trades from highest cap to lowest cap. That way after the market opens I can start trading the stocks at the top knowing that they will have a narrow bid/ask first. I can just glance at the bid/ask delta and mentally compare it to the price.
That setup enables me to quickly go down the trade tables clicking on the stocks, evaluating the chart, bid/ask delta, and the level 2 quotes. If a stock meets my criteria I submit the market order. If not, I skip it and go to the next one in the table. As I scan down the table, I see that the bid/ask of the below stocks are still too wide so I go back to the top of the table to the largest cap stock that I skipped in the first run through the table. I continue that process until I have completed all the trades. Most weeks it takes less than 20 minutes. Occasionally it will take 30 minutes or a little longer.
It’s been a long time since I have tried to trade more than 1% of the ADT of a micro cap stock and I no longer trade any stocks with an ADT < $200K. That way I hardly ever affect the price outside of the bid/ask range prior to placing the trade.
Denny
Denny - Re: 9:30 to 10:00 - is this eastern time?
Yes
Denny
Denny:
Thanks for the details on your trading set up. I see that keeping trades to 1% of ADT would really help with market orders. Only about 1/3 of my trades are under 1% of ADT. Another 1/4 are in the 1-1.99% of ADT range so market orders might work for them as well. That leaves 4/10 of my trades over 2% of ADT for which I’d continue using limit orders. Another alternative would be to split my trading over two days to cut my percentage of ADT in half. That would result in about 4/5 of my trades being 2% or less of ADT.
Right now my trade slippage is around 0.47% divided as follows:
… 0.21% from the bid-ask spread. (I currently do my trades in the afternoon when the bid-ask spread is lower than the first half hour)
… 0.10% for moved prices. About 1/2 of my trades do not move the prices, but the rest move the prices a bit. I generally quit buying at 0.5% movement, but when selling it has sometimes gone over 1%.
… 0.10% from not trying to buy at the open. (this ranges from 0.05% to 0.18% for my three portfolios). See note below for details.
… 0.06% for brokerage trading fees. About half with with IB at 0.5c/share and the other half with a flat rate broker.
Since the biggest source of slippage for my type of stocks is the bid-ask spread, I’ve been doing the majority of my trading in the afternoon when the spreads are the tightest.
By waiting to the early afternoon instead of the open, my estimate is about 0.10% of slippage is added. This is just an educated guess based on comparing sim results for trading at the open, at the average of the high and low, and at the close. If I were to trade at or near the close, the estimate slippage (vs the open) is around 0.15%.
What I don’t (yet) have a firm grasp on is precisely how much worse the bid-ask spread would be in the first 30 minutes and how much more (or less) I’d move the prices by trading at the open. If I had to guess at this point, I think I might save a bit (perhaps 0.05%-0.07%) by trading at the open, espeically if I split my trades over two days to have a lower percentage of ADT. Given the average turnover of my portfolios that would earn me 0.5% to 1.5% more in annual returns. By itself that’s not all that much. However, what intrigues me about Denny’s approach is the ability to get several trades at a better price than the open and thereby potentially improve the gain.
If anyone has been through similar process of deciding between trading at the open versus noon or some other time of day, I’d be most interested in what you’ve discovered.
Brian
Brian,
Others will have much more to offer. I did study of a fairly large number of trades with ADT(20) > $1,200,000 through Folio Investing (window trades). I got the slippage as 0.3% +/- 0.2% which was in line with variable slippage. This study compared the fill price to the opening as per Yahoo Finance. This sim for the port I am using shows bottom 20% to be $1,400,000 as there is no cut-off on market cap or liquidity.
Don’t know if this helps at all. Would be interested in the lower cut-off for liquidity of your universe.
Regards,
Jim
Brian,
I always trade MicroCaps in the early morning for a number of reasons. First, on average, there is higher liquidity in the morning then later in the day (except near the close). Second, the difference between my MicroCap Sims run at next open vs. next close is over 15%/year return, the average gain/stock is >0.5%, and the % Winners is >0.8%. So on average, the stocks increase in price during the day.
You might consider trading earlier in the day, if not right after the open. As long as you don’t give up an additional 0.5% /stock compared to trading in the afternoon you mighe come out ahead. Test the difference between next open and next close for your Sims to see if you get similar results.
Instead of breaking up your trades to trade the next day, consider trading those 10 to 30 minutes apart. I find that is enough time to not move the price if the bid/ask ratio is still small.
Denny
Jim:
Thanks for the data from your trades using Folio windows. I assume your 0.3% slippage was for using the morning window at Folio.
The bottom 20% ADT for my trades is $270,000.
Denny:
For my lower ADT stocks, I currently split my orders up into to 2 to 4 lots spread over 10 to 20 minutes. I’ll give it a try spreading them out more.
For my portfolios the difference in annual returns is smaller than yours for trading at the open versus the close (1%-4% versus 15%). These numbers assume that the bid-ask spread would get as narrow by 10 AM as it is by my current trading time of around 1 PM. I suspect the spread is larger at 10AM based on a detailed study I did several years ago. Perhaps I should redo the study since spread might be narrowing more quickly today.
Portfolio A
Annual Return: 50.1%/yr (open), 47.7% (H/L average), and 46.3 (close) so 2.4% and 3.8% is being given up.
Average Trades: 5.31% (open), 5.08% (avg HL), 4.99% (close) so giving up 0.25%-0.31% per trade with turnover of 9.3x
Portfolio B
Annual Return: 50.7%/yr (open), 49.0% (H/L average), and 47.8 (close) so 1.7% and 2.0% is being given up.
Average Trades: 6.47% (open), 6.28% (avg HL), 6.14% (close) so giving up 0.19%-0.33% per trade with turnover of 6.9x
Portfolio C
Annual Return: 43.5%/yr (open), 42.4% (H/L average), and 41.1 (close) so 1.1% and 2.4% is being given up.
Average Trades: 3.91% (open), 3.83% (avg HL), 3.74% (close) so giving up 0.08%-0.17% per trade with turnover of 11x
Brian