Ms. Carey, I appreciate your willingness to participate on these boards and address your critics. Others in the past have simply deleted their posts, taken their toys and run home to mommy (or their lawyer). I respect your point of view and your goal of giving your readers accurate information. But since you asked... I think your weightings are appropriate for traditional online brokers (and probably most of Barron's readers use these, so I have no problem with this category). In this category I would put most intermediate to long term investors. However, for "direct access" customers, which include most of those on this forum, the requirements for brokers are much different. The level of sophistication of a professional daytrader (or someone who aspires to this) is way beyond what is serviced by the traditional online brokers. For me, the only thing I want out of a broker is an efficient execution, and nothing more. This includes execution speed, cost, and user interface as it relates to quick order entry. I couldn't care less about the ability to research fundamentals or having access to other financial services. I can get all of that elsewhere if needed. In fact, the addition of these services will probably drive up costs and divert focus from my primary goal, efficient execution. I would have weighted direct access brokers this way: Execution Speed: 1.5 Cost: 1.25 Ease of Use: 1.25 Reports/Customer Access: 1.0 Range of Offerings: .5 Research Ammenities: .5 Just one guy's opinion...and I never get invited to any focus groups Regards, Kirk
Here's the gory detail on exactly what goes into the rankings. Merely saying the names of the various categories doesn't give the full picture. The following paragraphs are clipped from my March 12 feature. I apologize in advance for the length of this post, but I wanted to point out exactly what I look for when I concoct the rankings. It isn't a mysterious process -- it's spelled out, in detail, in print. (Please note the absence of "advertising in Barron's" or "kissing up to the reviewer.") The following is clipped from http://interactive.wsj.com/archive/retrieve.cgi?id=SB984177447765727969.djm&template=doclink.tmpl ------- snip ------ We assumed typical Barron's readers (if there are such persons) could avail themselves of benefits for high account-balance holders -- over $100,000 in assets -- and make three round-trip trades per month. We made significant changes to our rating system, evaluating the brokers' offerings and performance in six areas, assigning a point value of 0 (low) to 5 (high) in each category. We then weighted the results according to the presumed importance to investors, adding to two the measures, and halving another. What would it take to get the so-far unobtainable Five Stars? For starters, clearly written reports updated in real-time, plus price improvement on limit orders. Many readers wrote in to object to payment for order flow, so our rating system this year penalizes brokers who accept such payments. Proprietary research, well-organized links to third-party research, and great customer service also would be musts. We looked for a streamlined order entry process, minimizing (or eliminating) the number of verification screens by utilizing field-by-field checking during data entry, while maximizing the amount of data available at the time of the trade. A real-time quote, presented before the order is entered, is essential. Drop-down list boxes, or links to trading screens from a positions report or a research window, eliminating the need to enter ticker symbols, are a major plus. For example, if you select "Sell," there should be a drop-down box with the ticker symbols of issues in your portfolio, as well as the number of shares currently held, to avoid data-entry errors and inadvertent short sales. We'd like to see an easy-to-read verification screen that is more than an echo of the order entry screen; that spells out the name of the security being traded, and shows the total value of the trade, including commission charges. Final requests: an instant link to the status of the order, or a pop-up confirmation screen, followed by real-time updating of the trader's portfolio, plus portfolio analysis reports, with links to news and research, as well as transaction history going back at least 90 days. Readers also have emphasized the importance of tracking the tax consequences of each trade. We ranked the brokers in six key areas, awarding or subtracting points in numerous sub-categories: Trade Execution Process: Reader feedback inspired us to focus on the process of placing and confirming trades, which can't be ascertained by using a demo. We broke the points awarded down into six sub-categories: A real-time quote on the trading screen earns one point. Partial credit was given to brokers that display the real-time quote and then force a move to the trading screen. The quality of the trading screen itself was worth two points. A well-organized screen that helps the investor avoid data entry errors, and locates potential problems before they result in expensive effects, is key. Brokers offering price improvement for limit orders received 1/2 point. Brokers that accept payment for order flow lose 1/2 point. A pre-filled order entry screen was worth up to one point. We looked for the ability to click on a trading link while viewing a positions report, and have the order entry screen set up for the investor. The full point went to brokers who filled in the number of shares available in the investor's account when placing a Sell order. Easily accessible order status reports and pop-up confirmations were worth up to 1/2 point. We executed equity trades during market hours, performing a market buy and a limit sale of a Nasdaq stock. Following the market buy, we evaluated the execution and portfolio reports. After the limit sell, we examined the open order reports and looked at ways for the trader to follow the progress of the order. We also examined the mutual fund and options entry screens, though we did not place orders there. A "5" in this category means the order entry and execution process flowed easily from one step to the next, with real-time information available when needed. The availability of price improvement strategies and the absence of payment for order flow are necessary to attain a "5." We added 25% to this score to emphasize its importance to Barron's readers. Ease of Use: How easy was it to navigate around the site? Does the layout of the site make sense and minimize the number of mouse clicks it can take to get from one place to another? Sites that keep you just a mouse click or two away from any other area get higher rankings than those that require navigation through an online obstacle course. A "5" in this category means the site was easy to use, well designed, and doesn't bog down when moving from screen to screen. Range of Offerings: In past years, we've combined this category with others, but readers have prompted us to give it its own place in the sun again. We awarded points for equities that can be traded online, with partial points given for those that can be traded with the broker, but only offline. Points were given for the ability to trade stocks long and short, mutual funds, bonds, simple options and complex options. A fraction of a point was awarded for other online investing opportunities, including CDs and banking services. A "5" in this category means you can execute all of these transactions online. Research Amenities: This category measures the quality and accessibility of proprietary research, third-party research, quotes and charting. We assigned up to one point for proprietary research available only to account holders, also determining whether additional goodies are given to those with large accounts. Up to two points were awarded, depending on the third-party research available and how well it was integrated into the brokerage's site. Quotes and charts earned up to two points also, with the higher rating going to brokers with access to real-time streaming quotes, powerful charting capabilities, and Level II quote accessibility. The brokers who received the highest points in this category gave exceptional research and quote services to high-account-balance holders. Reports and Customer Access: This category measures the quality of portfolio analysis reporting (up to one point); the quality of tax reports available, including how far into the past transaction history is posted (up to 1 1/2 points); and the availability and quality of online and offline help (up to two points). Brokers received 1/2 point for having wireless trading access. Online help includes live chat capability, frequently asked question files, and investor education. Offline help was assessed through the assistance of a group of friends, who made random calls to customer service asking relatively intelligent questions. (I've found that certain brokers tend to be extra nice to anyone who calls up claiming to be Theresa Carey.) We added 25% to the score in this category, emphasizing what Barron's readers want: How much am I worth, what are the tax consequences of my trades, and what can you do for me right now? Costs: This year, we looked at commissions for stock and options trades, and margin interest rates. Higher points are assigned to lower costs. We looked at the cost of three round-trip stock trades, assuming one side of the trade was at market and one side was a limit order. Up to three points were awarded for stock commissions. We also looked at the cost of trading 10 options contracts, with one point going out in this category. Last but not least, up to one point could be earned for lower margin interest rates on balances of $25,000 and $50,000. A "5" in this category could be earned by a broker with commissions of under $50 total for three round-trips, commissions of under $50 for 10 options contracts, and margin interest rates under 7.5%. We reduced the weight of costs in the overall ranking by 50% to reflect the importance of trade quality and service over price. ------- snip ------ So that's the nitty-gritty of what goes into the list of numbers that started this thread. Thanks for listening. -- T.
"We assumed typical Barron's readers (if there are such persons) could avail themselves of benefits for high account-balance holders -- over $100,000 in assets -- and make three round-trip trades per month" My initial thought is that these people shouldn't be looking at any of the direct access brokers you mention in the article except Schwab. 3 r/t's a month hardly requires the typical platforms offered by direct access brokers. I'm sure the rest of the national discount firms(waterhouse,etrade, ameritrade) will soon offer some kind of watered down version of direct access execution. This will satisfy your readers need for improved execution while maintaining the level of services they have become accustom to in dealing with these firms over the years.
Dear Ms. Carey, Two years in Japan, eh? I was two years in China. I miss it. You asked for suggestions for weightings. I'd like to offer a slightly different approach. How about rethinking the research design, moving from a composite index approach (various point totals in various qualitative subcategories) to an approach based on a combination of various investor/trader segments' observed choices (quantitative), and their reported reasons for those choices (qualitative). While I'd want to do a lot more thinking about the research design of each, here is an off-the-top-of-my head idea about the quantitative side. For starters, segment the universe of traders/investors by two or three dimensions: number of trades per month, account size, and, perhaps, average duration per trade (though my guess is that this would correlate inversely with number of trades per month). Maybe you wind up with a 3x3 table, using High, Middle, Low on each of two dimensions. Then observe the "stickiness" of brokers within each segment, to see how the members of that segment are voting with their accounts. Let me explain three ways of possibly measuring stickiness, and then explain why I would slightly prefer that variable to "number of accounts using a particular broker platform." Stickiness: The duration an investor/trader keeps his/her account with a particular broker platform as a percentage of the duration he/she has been a member of that investor/trader segment. You might want to create a second variable that breaks out the same data for the subset of that segment that has changed broker platforms at least once, since those people would have had a greater personal basis upon which to compare. For that matter, you might want to create still a third variable that measures duration not as a percentage (treating newbies and oldtimers equally) but as actual duration (which would give greatest weight to the choices of those people who have been members of that segment the longest). For the first of these stickiness measures, let's say an active trader over a four-year period tries Platform A for a year and half, then moves to Platform B for a month, and then to Platform C for the next 35 months. The scores: Platform A gets 18/48ths, B gets 1/48th, and C gets 29/48ths. For the second measure of stickiness, Platform A would get no score (since there was no comparative basis to precede it), Platform B gets 1/30th, and Platform C gets 29/30ths. And so on. Again, this is off the top of my head, but the point is to try to measure how people are actually voting with their feet. My impression is that among active traders there is a large and dedicated following of the RealTick platform and the various brokers that service it, another bunch happy with CyberTrader, and a growing following of Interactive Brokers. Incidentally, I said I'd explain why I'd prefer a measure of stickiness to the number of accounts using a particular broker platform: I think the latter tends to bias the results according to which broker platform has been most heavily advertised to members entering that investor/trader segment, and penalizes those with low advertising budgets. As for the qualitative side, you could ask members of each segment why they've stayed or why they've moved, and see which variables each segment seems to weight most heavily. Let me acknowledge that this would be a big undertaking, especially so in the first year. But perhaps you could find a group of underwriters. Here's ambition: perhaps even develop yourself into the J.D. Powers of vendors serving the investing/trading public. I've gone on long enough. Thanks for your patience. Regards, NT
NT, Thank you very much for your support and writing in such an eloquent manner. Ms. Carey, I as well appreciate your involvement on this board. Assuming I understand things correctly, I'd like to point out one area of your testing that may bias one broker against another by mere luck. You state that you sent in one buy and one sell (limit and market). With NASDAQ stocks if a market maker is showing the best price and an order gets routed to them, they often exercise the right to hold onto the order for upto 30 seconds. Thus if you send a market order via broker A and it gets sent to ISLAND, execution should be immeditate. However, if Broker B's order gets sent to a non-ECN (market maker), the fill could take 1/2 a minute. This should improve as SUPERSOES gets implemented. In any event, maybe you already took this into consideration. If not, next time around it may be appropriate (if not too costly) to route and compare many orders via all avenues in order to gain an effective statistical sample.
Pathetic........... ``This past quarter marks the completion of our transformation from a software developer to a brokerage firm that targets institutional and active traders,'' said Bill Cruz, Co-Chairman of the Board and Co-CEO. ``It has been the most significant quarter in the 19-year history of our company. We believe our selection by Barron's as the top-ranked direct-access brokerage firm ONLY DAYS AFTER OUR LAUNCH is an early indication of the great success this company will achieve in 2002 and beyond.'' taken from Tradestation quarterly results statement.
Looks like Barrons listened. IB rated second only to TN in trade execution in latest survey of online brokers. Although they didn't include IB in their "What It Costs, What You Get" section. Not sure why.
"Although they didn't include IB in their "What It Costs, What You Get" section. Not sure why." I emailed the reporter and she responded that she's not sure how it happened. IB was included in that table in the manuscript and draft she received. She also noticed it when she looked at the print edition. She said: "I hope to get my pals in the art department to run the Interactive Brokers line next week." 99