WEBVTT 00:00:00.000 --> 00:00:01.940 Welcome to the Deep Dive, where we take your 00:00:01.940 --> 00:00:05.139 stack of sources and distill the absolute essence. 00:00:05.860 --> 00:00:08.519 Today's mission is all about knowledge optimization. 00:00:09.300 --> 00:00:12.160 We're giving you the shortcut to being well informed 00:00:12.160 --> 00:00:15.199 on two really critical financial metrics. That's 00:00:15.199 --> 00:00:18.199 right. Metrics that... Every investor and honestly, 00:00:18.339 --> 00:00:20.940 even anyone interested in organizational efficiency 00:00:20.940 --> 00:00:24.059 really needs to understand. We are diving deep 00:00:24.059 --> 00:00:26.719 into the expense ratio and the Sharpe ratio. 00:00:26.940 --> 00:00:29.359 And these are not just, you know, abstract concepts 00:00:29.359 --> 00:00:31.899 for Wall Street analysts. If you, the learner, 00:00:32.060 --> 00:00:35.100 are evaluating a retirement fund or maybe you're 00:00:35.100 --> 00:00:37.380 vetting a hedge fund manager or even just deciding 00:00:37.380 --> 00:00:40.000 which charity to support. Right. These two ratios 00:00:40.000 --> 00:00:42.460 are the lenses. They help you evaluate both the 00:00:42.460 --> 00:00:45.439 guaranteed cost of accessing a return and, well, 00:00:45.539 --> 00:00:47.820 the. quality of that return, adjusting for the 00:00:47.820 --> 00:00:49.000 risk you're taking. And we're going to spend 00:00:49.000 --> 00:00:51.100 time on the surprising implications, not just 00:00:51.100 --> 00:00:53.119 the textbook definition. Absolutely. Okay, let's 00:00:53.119 --> 00:00:55.140 unpack this. We have some excellent source material 00:00:55.140 --> 00:00:58.780 here defining both of these terms and exploring 00:00:58.780 --> 00:01:01.420 their implications in mutual funds, hedge funds, 00:01:01.560 --> 00:01:04.299 and even charity organizations. It's a wide -ranging 00:01:04.299 --> 00:01:07.040 topic. It is. So let's start with the metric 00:01:07.040 --> 00:01:09.299 that's, well, it's concrete and it's guaranteed 00:01:09.299 --> 00:01:12.019 because it silently eats away at returns every 00:01:12.019 --> 00:01:15.090 single year. The expense ratio. Right. The invisible 00:01:15.090 --> 00:01:17.629 cost. The invisible cost of investing. So the 00:01:17.629 --> 00:01:20.269 expense ratio, or ER, is actually pretty straightforward 00:01:20.269 --> 00:01:22.390 in what it represents. It's the total percentage 00:01:22.390 --> 00:01:24.569 of a fund's assets that gets consumed annually 00:01:24.569 --> 00:01:27.390 by all the operational costs of just running 00:01:27.390 --> 00:01:30.129 that fund. Okay. We're assuming our listener 00:01:30.129 --> 00:01:31.909 knows the basics, so let's jump right into the 00:01:31.909 --> 00:01:35.090 nuance. The ER is that total percentage for administrative, 00:01:35.450 --> 00:01:38.989 management, all those operating expenses. What's 00:01:38.989 --> 00:01:41.439 the key implication of that percentage hit? The 00:01:41.439 --> 00:01:45.019 implication is, well, it's brutal and direct. 00:01:45.159 --> 00:01:48.980 If a fund has an ER of, say, 1 .2 percent, then 00:01:48.980 --> 00:01:51.620 1 .2 percent of the fund's total assets are just 00:01:51.620 --> 00:01:54.260 subtracted annually. Regardless of performance. 00:01:54.560 --> 00:01:56.500 Regardless of whether the fund makes money or 00:01:56.500 --> 00:01:59.219 loses money. It's a structural drag that directly 00:01:59.219 --> 00:02:01.760 reduces your gross return before you see a single 00:02:01.760 --> 00:02:04.480 cent of profit. Now, a common misconception is 00:02:04.480 --> 00:02:06.599 that the expense ratio represents the entire 00:02:06.599 --> 00:02:08.620 cost of owning that fund. But that's misleading, 00:02:08.780 --> 00:02:11.759 isn't it? Oh, it is absolutely crucial to be 00:02:11.759 --> 00:02:14.900 clear on what the ER does not include. It doesn't 00:02:14.900 --> 00:02:17.719 cover sales loads. Which are the upfront or deferred 00:02:17.719 --> 00:02:20.460 charges. Exactly. The sales charges you pay when 00:02:20.460 --> 00:02:23.259 you buy or sell shares. And it also doesn't cover 00:02:23.259 --> 00:02:25.199 brokerage commissions. I mean, those are the 00:02:25.199 --> 00:02:27.840 transaction costs the fund manager racks up when 00:02:27.840 --> 00:02:29.800 they're buying and selling securities inside 00:02:29.800 --> 00:02:32.340 the fund. So those are separate costs borne by 00:02:32.340 --> 00:02:34.639 the investor. They are, and they can be substantial, 00:02:34.900 --> 00:02:36.659 especially for funds that are really actively 00:02:36.659 --> 00:02:39.800 traded. So the ER is like the internal engine's 00:02:39.800 --> 00:02:42.460 operating cost. And that cost isn't just a random 00:02:42.460 --> 00:02:45.360 number. It's influenced by some pretty clear 00:02:45.360 --> 00:02:48.020 structural factors. What drives it up or down? 00:02:48.219 --> 00:02:50.939 There are two primary factors. First one is fund 00:02:50.939 --> 00:02:53.520 size. This all comes down to economies of scale. 00:02:53.719 --> 00:02:56.659 Okay. Think about it. A smaller fund has the 00:02:56.659 --> 00:03:00.000 same mandatory fixed costs as a huge fund. Things 00:03:00.000 --> 00:03:02.800 like annual regulatory audits, legal compliance, 00:03:03.120 --> 00:03:05.599 even rent for the office. If you have to spread 00:03:05.599 --> 00:03:08.180 those fixed costs over a small asset base, the 00:03:08.180 --> 00:03:11.340 resulting ER as a percentage just has to be higher. 00:03:11.500 --> 00:03:13.800 So as the fund gets bigger, that percentage should 00:03:13.800 --> 00:03:16.520 drop. It should theoretically drop. But as we're 00:03:16.520 --> 00:03:18.860 going to discuss in a bit, that often doesn't 00:03:18.860 --> 00:03:20.900 happen as much as you think because of how the 00:03:20.900 --> 00:03:23.020 other fees are structured. And the second factor 00:03:23.020 --> 00:03:25.090 has to do with effort. Yeah. Management style. 00:03:25.169 --> 00:03:28.229 Yes. So passively managed funds, you know, the 00:03:28.229 --> 00:03:30.289 ones just tracking an index like the S &P 500, 00:03:30.509 --> 00:03:33.569 they typically have significantly lower expense 00:03:33.569 --> 00:03:35.770 ratios. We're talking basis points here, like 00:03:35.770 --> 00:03:39.550 hundreds of a percent, maybe 0 .05 percent or 00:03:39.550 --> 00:03:42.210 0 .10 percent. Because they're not doing a ton 00:03:42.210 --> 00:03:44.409 of research. Right. They require minimal research, 00:03:44.569 --> 00:03:46.969 fewer decision making resources. On the flip 00:03:46.969 --> 00:03:49.530 side, you have actively managed funds. This is 00:03:49.530 --> 00:03:52.090 where a highly paid team of analysts and portfolio 00:03:52.090 --> 00:03:54.860 managers is trying. to beat the market. That 00:03:54.860 --> 00:03:57.719 requires enormous overhead and research spending, 00:03:57.840 --> 00:04:00.759 which drives their ERs way higher, often into 00:04:00.759 --> 00:04:04.340 the 1 .0 % to 2 .0 % range. Let's focus on a 00:04:04.340 --> 00:04:06.539 specific and often pretty controversial line 00:04:06.539 --> 00:04:09.879 item in U .S. funds, the 12B1 fee. Our source 00:04:09.879 --> 00:04:12.300 details the specific caps on this, but instead 00:04:12.300 --> 00:04:13.960 of just reciting the rule, let's talk about the 00:04:13.960 --> 00:04:16.639 implication. The 12B1 fee is maybe the most fascinating 00:04:16.639 --> 00:04:18.879 component because it is essentially the fund's 00:04:18.879 --> 00:04:21.100 marketing and distribution expense, and it's 00:04:21.100 --> 00:04:23.600 paid for by you, the investor. My money is paying 00:04:23.600 --> 00:04:25.579 for the ads. Your capital is paying to attract 00:04:25.579 --> 00:04:28.759 other investors. This fee, it's generally capped 00:04:28.759 --> 00:04:33.319 by Fenar at 1 .00%, broken down as 0 .75 % for 00:04:33.319 --> 00:04:36.540 distribution and 0 .25 % for shareholder servicing. 00:04:36.680 --> 00:04:39.199 That feels really counterintuitive. I'm paying 00:04:39.199 --> 00:04:42.259 a fee so the fund can grow. Which you could argue 00:04:42.259 --> 00:04:44.459 benefits the fund manager in the company more 00:04:44.459 --> 00:04:46.459 than it benefits me, the existing shareholder. 00:04:46.600 --> 00:04:48.319 And that's precisely the implication. I mean, 00:04:48.339 --> 00:04:50.579 proponents will argue that growth ultimately 00:04:50.579 --> 00:04:52.879 benefits shareholders by helping the fund achieve 00:04:52.879 --> 00:04:55.439 those economies of scale we just mentioned. Right. 00:04:55.579 --> 00:04:58.180 But in practice, it's a direct cost for marketing. 00:04:58.319 --> 00:05:01.220 For a well -informed investor, if you see a substantial 00:05:01.220 --> 00:05:04.639 12B1 fee, it should raise a flag about where 00:05:04.639 --> 00:05:06.480 the fund's priorities are. Is it performance 00:05:06.480 --> 00:05:09.259 or is it expansion? Now, here's where the analysis 00:05:09.259 --> 00:05:11.740 moves beyond just... the definition and into 00:05:11.740 --> 00:05:14.959 actual strategy. Unlike trying to predict future 00:05:14.959 --> 00:05:18.279 stock returns, which is, well, it's notoriously 00:05:18.279 --> 00:05:20.480 difficult. The fool's errand, some would say. 00:05:20.639 --> 00:05:23.500 Fund expenses are highly predictable. This is 00:05:23.500 --> 00:05:25.920 a huge advantage for you as an investor. Funds 00:05:25.920 --> 00:05:29.300 with... Historically high expense ratios tend 00:05:29.300 --> 00:05:32.459 to, you know, continue having high ratios. This 00:05:32.459 --> 00:05:35.060 is why our source material really stresses examining 00:05:35.060 --> 00:05:37.980 the financial highlights section in the prospectus. 00:05:38.100 --> 00:05:40.579 Look at the ER history over the last five years. 00:05:40.680 --> 00:05:43.180 If it hasn't dropped, there's very little reason 00:05:43.180 --> 00:05:45.199 to believe it suddenly will. And the question 00:05:45.199 --> 00:05:48.910 is why. Why is it so hard for a fund that's been 00:05:48.910 --> 00:05:52.149 around for years to suddenly slash its expenses? 00:05:52.589 --> 00:05:55.129 Doesn't the massive growth of assets under management, 00:05:55.290 --> 00:05:58.529 the AUM, inherently make that ratio drop? That 00:05:58.529 --> 00:06:00.970 is the critical question. And it all comes down 00:06:00.970 --> 00:06:03.829 to the mix of fixed and variable costs because 00:06:03.829 --> 00:06:05.589 they behave very differently when you measure 00:06:05.589 --> 00:06:07.389 them as a percentage. It requires some careful 00:06:07.389 --> 00:06:09.680 differentiation. Okay, let's try to use that 00:06:09.680 --> 00:06:12.160 small business analogy you suggested to simplify 00:06:12.160 --> 00:06:14.560 this because it can be a conceptual hurdle. Let's 00:06:14.560 --> 00:06:16.860 break down the costs. The majority of the biggest 00:06:16.860 --> 00:06:19.180 operational costs in a mutual fund, like the 00:06:19.180 --> 00:06:21.699 management fee, that's set as a percentage of 00:06:21.699 --> 00:06:23.870 AUM. So let's say, for example, the management 00:06:23.870 --> 00:06:28.389 fee is 0 .75%. That 0 .75 % is fixed on a percentage 00:06:28.389 --> 00:06:31.810 basis. If the fund doubles in size, the manager's 00:06:31.810 --> 00:06:34.250 dollar payout doubles, sure, but the percentage 00:06:34.250 --> 00:06:37.870 taken from your assets stays at 0 .75%. So it's 00:06:37.870 --> 00:06:39.889 a variable dollar cost but a fixed percentage 00:06:39.889 --> 00:06:42.350 cost. Exactly. And that percentage is sticky. 00:06:42.449 --> 00:06:45.069 It generally doesn't change unless the AUM hits 00:06:45.069 --> 00:06:47.189 some pre -agreed breakpoint, which is pretty 00:06:47.189 --> 00:06:49.430 rare for older funds. So the biggest piece of 00:06:49.430 --> 00:06:52.370 the pie... the manager's compensation is mathematically 00:06:52.370 --> 00:06:55.339 designed to stay constant. As a percentage drag 00:06:55.339 --> 00:06:58.000 on me, the investor. Precisely. Now you compare 00:06:58.000 --> 00:07:00.420 that to the truly fixed costs, the costs that 00:07:00.420 --> 00:07:03.160 are a fixed lump sum of dollars. Like rent. Rent, 00:07:03.220 --> 00:07:05.600 the annual audit fee, regulatory filing costs. 00:07:06.019 --> 00:07:08.939 If the fund is small, maybe $100 ,000 audit fee 00:07:08.939 --> 00:07:12.120 consumes 0 .25 % of AUM. But if that fund grows 00:07:12.120 --> 00:07:15.519 tenfold, that same $100 ,000 audit fee now only 00:07:15.519 --> 00:07:19.500 consumes 0 .025 % of AUM. This is where economies 00:07:19.500 --> 00:07:21.639 of scale actually provide a benefit. I see. So 00:07:21.639 --> 00:07:23.519 while those fixed lump sum costs, their percentage 00:07:23.519 --> 00:07:25.230 of... impact goes down as the fund grows, the 00:07:25.230 --> 00:07:27.290 majority of the cost, that big management fee, 00:07:27.370 --> 00:07:29.490 it stays glued to the total assets. Correct. 00:07:30.160 --> 00:07:32.740 And because that management fee usually just 00:07:32.740 --> 00:07:35.800 dwarfs the fixed dollar costs, the overall expense 00:07:35.800 --> 00:07:38.639 ratio tends to be pretty constant. That makes 00:07:38.639 --> 00:07:41.379 significant future reductions very unlikely for 00:07:41.379 --> 00:07:44.100 historical funds. If you buy a fund with a 1 00:07:44.100 --> 00:07:47.500 .5 % ER today, you are essentially signing up 00:07:47.500 --> 00:07:50.920 for a permanent 1 .5 % headwind against your 00:07:50.920 --> 00:07:53.060 returns for the lifetime of that investment. 00:07:53.360 --> 00:07:55.779 Now, let's pivot to a fascinating implication 00:07:55.779 --> 00:07:59.240 of this cost structure. The expense ratios damage 00:07:59.240 --> 00:08:01.800 isn't linear. It's disproportionately devastating 00:08:01.800 --> 00:08:03.620 depending on the type of investment you choose. 00:08:03.839 --> 00:08:06.040 And here's where the math really moves from just 00:08:06.040 --> 00:08:08.600 accounting into investment strategy. We need 00:08:08.600 --> 00:08:10.920 to look at the ER relative to the expected gross 00:08:10.920 --> 00:08:13.220 returns of the asset class. So let's use three 00:08:13.220 --> 00:08:15.220 categories with dramatically different expected 00:08:15.220 --> 00:08:17.480 returns. Equity funds, bond funds, and money 00:08:17.480 --> 00:08:19.680 market funds. And we'll use a high fixed one 00:08:19.680 --> 00:08:22.399 per turn ER across the board just for this illustration, 00:08:22.600 --> 00:08:24.300 even though, you know, real world index funds 00:08:24.300 --> 00:08:26.079 are much cheaper. So starting with the highest 00:08:26.079 --> 00:08:28.629 return environment, equity. funds. Historically, 00:08:28.649 --> 00:08:31.370 let's assume an average gross return of 9 % for 00:08:31.370 --> 00:08:33.549 a typical equity fund before you take out any 00:08:33.549 --> 00:08:37.590 fees. That 1 % ER consumes approximately 11 % 00:08:37.590 --> 00:08:40.230 of your return. That's 1 divided by 9. It's a 00:08:40.230 --> 00:08:42.490 significant cut. It is, but you still retain 00:08:42.490 --> 00:08:45.230 89 % of the gross return. Okay, now we move to 00:08:45.230 --> 00:08:47.830 bond funds, a medium return environment. Bond 00:08:47.830 --> 00:08:50.870 funds typically have lower historical gross returns, 00:08:50.889 --> 00:08:54.840 so let's say 8%. That same 1 % ER now consumes 00:08:54.840 --> 00:08:58.480 about 12 .5 % of the investor's return, 1 divided 00:08:58.480 --> 00:09:01.220 by 8. The fee itself hasn't changed, but the 00:09:01.220 --> 00:09:03.840 relative drag is already higher because the denominator 00:09:03.840 --> 00:09:06.500 of the return is smaller. And this effect becomes 00:09:06.500 --> 00:09:09.279 extreme. when we look at those low -yield, short 00:09:09.279 --> 00:09:11.419 -term investments like money market funds? This 00:09:11.419 --> 00:09:13.500 is where the expense ratio becomes truly devastating. 00:09:13.899 --> 00:09:16.200 Assuming a typical long -term historical gross 00:09:16.200 --> 00:09:20.379 return of only 5%, that 1 % ER now consumes a 00:09:20.379 --> 00:09:22.639 substantial 20 % of the historical total return. 00:09:22.759 --> 00:09:25.379 That's 1 divided by 5. Wow. So in this simplified 00:09:25.379 --> 00:09:28.580 example, the same 1 % fee is nearly twice as 00:09:28.580 --> 00:09:30.360 painful in a bond fund as it is in an equity 00:09:30.360 --> 00:09:32.440 fund, and then almost double again in a money 00:09:32.440 --> 00:09:35.090 market fund. Exactly. And this analysis becomes 00:09:35.090 --> 00:09:36.929 even more critical when you factor in inflation. 00:09:37.330 --> 00:09:39.710 If you have a money market fund that's yielding 00:09:39.710 --> 00:09:43.950 5 % gross and a 1 % ER, your net return is 4%. 00:09:43.950 --> 00:09:47.289 If inflation is 3%, your real return is only 00:09:47.289 --> 00:09:51.110 1%. So that 1 % fee consumed a massive chunk 00:09:51.110 --> 00:09:53.690 of the profit and a little bit of inflation ate 00:09:53.690 --> 00:09:56.179 up most of what was left. That's right. The key 00:09:56.179 --> 00:09:58.179 takeaway here for you, the listener, is clear. 00:09:58.480 --> 00:10:01.379 You have to consider the ER relative to the expected 00:10:01.379 --> 00:10:03.919 returns of the asset class. In the current environment, 00:10:04.039 --> 00:10:06.200 where fixed income returns have often been compressed 00:10:06.200 --> 00:10:09.279 or even negative in real terms, a high ER is 00:10:09.279 --> 00:10:11.279 far more damaging to a low return investment 00:10:11.279 --> 00:10:13.879 than to a potentially high return, high volatility 00:10:13.879 --> 00:10:16.500 one. Before we move on from fund analysis, we 00:10:16.500 --> 00:10:18.360 need to touch on a specific issue that can often 00:10:18.360 --> 00:10:20.759 cloak the true expense structure, particularly 00:10:20.759 --> 00:10:22.720 in funds that are just getting started, this 00:10:22.720 --> 00:10:25.929 concept of waivers and recoupments. is so essential 00:10:25.929 --> 00:10:27.610 for investors who are looking at newer funds. 00:10:27.909 --> 00:10:30.570 As we mentioned, new funds inherently have high 00:10:30.570 --> 00:10:33.490 expenses because their asset base is small and 00:10:33.490 --> 00:10:37.269 those fixed lump sums, they hit hard. So to attract 00:10:37.269 --> 00:10:39.750 early investors, the fund advisors will often 00:10:39.750 --> 00:10:42.409 enter into what are called waivers or reimbursement 00:10:42.409 --> 00:10:45.070 agreements. So the advisor is temporarily footing 00:10:45.070 --> 00:10:48.259 the bill. to artificially push down the reported 00:10:48.259 --> 00:10:51.600 ER just to make the fund look competitive. What's 00:10:51.600 --> 00:10:53.559 the catch? The catch is the recoupment plan. 00:10:53.840 --> 00:10:56.500 These waived or reimbursed amounts, the money 00:10:56.500 --> 00:10:58.720 the advisor essentially lent the fund, often 00:10:58.720 --> 00:11:02.220 have to be repaid by the fund later on. SEC rules 00:11:02.220 --> 00:11:04.779 generally allow funds up to three years from 00:11:04.779 --> 00:11:06.539 the date the expense was incurred to collect 00:11:06.539 --> 00:11:08.659 this money back. So an investor buys a fund in 00:11:08.659 --> 00:11:11.539 year one because the ER is, say, 0 .5 % because 00:11:11.539 --> 00:11:13.940 of a waiver, but that 0 .5 % was artificially 00:11:13.940 --> 00:11:16.450 low. Then in year three, the fund performs well, 00:11:16.549 --> 00:11:18.450 and suddenly the management is allowed to claw 00:11:18.450 --> 00:11:20.269 back those waived expenses from the previous 00:11:20.269 --> 00:11:23.710 two years. Precisely. The implication for future 00:11:23.710 --> 00:11:26.370 shareholders is that you may be required to absorb 00:11:26.370 --> 00:11:29.110 expenses that were incurred by the fund during 00:11:29.110 --> 00:11:31.970 prior years. It's like a hidden debt that subsequent 00:11:31.970 --> 00:11:34.309 investors are responsible for repaying. It feels 00:11:34.309 --> 00:11:36.899 like an ethical gray area. It is. The fiduciary 00:11:36.899 --> 00:11:40.059 incentive is to waive costs to attract assets, 00:11:40.299 --> 00:11:42.559 knowing that those costs can be recovered later 00:11:42.559 --> 00:11:45.059 from what will hopefully be a larger pool of 00:11:45.059 --> 00:11:47.259 shareholders. So the lesson here is that you 00:11:47.259 --> 00:11:49.340 can never just look at the current ER of a new 00:11:49.340 --> 00:11:52.360 fund. You have to check the prospectus for any 00:11:52.360 --> 00:11:54.779 active recoupment plan because that beautiful 00:11:54.779 --> 00:11:58.539 low ER might be hiding a future increase just 00:11:58.539 --> 00:12:00.659 waiting to be triggered. That's the key. We've 00:12:00.659 --> 00:12:02.820 seen the ER as a measure of investment leakage. 00:12:03.399 --> 00:12:05.960 But this concept of operational efficiency, how 00:12:05.960 --> 00:12:07.419 much money is consumed by the administrative 00:12:07.419 --> 00:12:10.620 engine, is essential in other sectors too. Let's 00:12:10.620 --> 00:12:12.820 look at how donors use a parallel metric in the 00:12:12.820 --> 00:12:15.320 world of nonprofits. This is a direct conceptual 00:12:15.320 --> 00:12:17.980 translation. For charities and for donors, the 00:12:17.980 --> 00:12:21.200 primary focus is the program expense ratio. It's 00:12:21.200 --> 00:12:23.039 the percentage of a charity's total expenses 00:12:23.039 --> 00:12:25.460 that go directly into executing the charitable 00:12:25.460 --> 00:12:27.879 mission, you know, the program services. So if 00:12:27.879 --> 00:12:31.759 a charity spends $100 total, And $85 of that 00:12:31.759 --> 00:12:34.879 goes to, say, feeding the hungry. The program 00:12:34.879 --> 00:12:38.320 expense ratio is 85%. Correct. Which leaves us 00:12:38.320 --> 00:12:41.059 with the remainder. The support service expense 00:12:41.059 --> 00:12:44.580 ratio, which is commonly, though often inaccurately, 00:12:44.600 --> 00:12:47.379 just called overhead. By definition, the program 00:12:47.379 --> 00:12:49.320 expense ratio plus the support service expense 00:12:49.320 --> 00:12:52.649 ratio, they have to equal 100%. And the simple 00:12:52.649 --> 00:12:54.909 common rubric is that a higher program expense 00:12:54.909 --> 00:12:58.149 ratio and thus lower overhead signals greater 00:12:58.149 --> 00:13:00.529 efficiency, which makes the charity more attractive 00:13:00.529 --> 00:13:03.009 to donors. That is the initial surface level 00:13:03.009 --> 00:13:05.570 assessment. But leading sources of information 00:13:05.570 --> 00:13:08.269 about charities like Charity Navigator and GuideStar, 00:13:08.409 --> 00:13:11.190 they strongly caution against blindly pursuing 00:13:11.190 --> 00:13:13.750 the lowest possible overhead. This is really 00:13:13.750 --> 00:13:16.110 where critical thinking is required. Why is optimizing 00:13:16.110 --> 00:13:18.690 for low overhead potentially a bad thing? Well, 00:13:18.769 --> 00:13:20.590 it can lead to what's called the starvation cycle. 00:13:21.200 --> 00:13:24.120 If a charity focuses obsessively on keeping its 00:13:24.120 --> 00:13:27.360 overhead staff, salaries, technology, facilities, 00:13:27.879 --> 00:13:31.779 fundraising below, say, 10%, they might underpay 00:13:31.779 --> 00:13:34.759 critical staff, use outdated and inefficient 00:13:34.759 --> 00:13:38.080 technology, and they might fail to invest adequately 00:13:38.080 --> 00:13:40.860 in effective fundraising or proper impact evaluation. 00:13:41.769 --> 00:13:43.669 So they look efficient on paper, but they're 00:13:43.669 --> 00:13:47.110 struggling to scale or maximize their long -term 00:13:47.110 --> 00:13:49.529 mission effectiveness. Precisely. They might 00:13:49.529 --> 00:13:52.350 have a fantastic program expense ratio, but they're 00:13:52.350 --> 00:13:54.710 inefficient at their core mission delivery because 00:13:54.710 --> 00:13:56.710 they haven't invested in the proper infrastructure. 00:13:57.129 --> 00:13:59.509 High efficiency requires professional management, 00:13:59.809 --> 00:14:01.750 good systems, the ability to measure results, 00:14:02.049 --> 00:14:04.169 and those things cost money. Donors need to look 00:14:04.169 --> 00:14:06.250 beyond that overhead percentage. They do. They 00:14:06.250 --> 00:14:08.470 need to look at factors like transparency, governance, 00:14:08.669 --> 00:14:11.110 strong leadership, and verifiable long -term 00:14:11.110 --> 00:14:12.710 outcomes. term results. Just to give our listener 00:14:12.710 --> 00:14:14.590 a frame of reference, what are typical overhead 00:14:14.590 --> 00:14:17.190 levels in the U .S. nonprofit sector? Based on 00:14:17.190 --> 00:14:19.889 2009 data from Charity Navigator, the national 00:14:19.889 --> 00:14:22.169 median for the support service expense ratio, 00:14:22.389 --> 00:14:25.629 the overhead, was 10%. But more importantly, 00:14:25.789 --> 00:14:27.690 the source noted that more than three -fourths 00:14:27.690 --> 00:14:30.190 of ranked charities had an expense ratio of less 00:14:30.190 --> 00:14:33.409 than 30%. If you see a charity with an overhead 00:14:33.409 --> 00:14:37.190 far exceeding 30 -35%, that is usually a legitimate 00:14:37.190 --> 00:14:39.919 red flag that needs some scrutiny. OK, before 00:14:39.919 --> 00:14:41.960 we make our grand transition to the Sharpe ratio, 00:14:42.279 --> 00:14:44.740 let's quickly anchor this expense ratio concept 00:14:44.740 --> 00:14:47.679 back in general business operations. Absolutely. 00:14:48.159 --> 00:14:50.700 Beyond funds and nonprofits, the expense ratio 00:14:50.700 --> 00:14:53.779 is a staple tool for finance and accounting professionals. 00:14:54.220 --> 00:14:56.820 In a business context, it just shows the percentage 00:14:56.820 --> 00:14:59.159 of an operation's gross revenues that is allocated 00:14:59.159 --> 00:15:01.600 to its running expenses. It sounds a bit like 00:15:01.600 --> 00:15:04.220 gross margin. But more detailed on the operating 00:15:04.220 --> 00:15:06.460 side. It's a measure of operational leakage and 00:15:06.460 --> 00:15:09.179 management effectiveness. Business managers use 00:15:09.179 --> 00:15:11.139 this index in their profit and loss statements 00:15:11.139 --> 00:15:14.539 to draft business plans, produce forecasts, and, 00:15:14.639 --> 00:15:17.059 you know, determine opportunities for cost cutting 00:15:17.059 --> 00:15:19.759 or revenue maximization. So it's an early warning 00:15:19.759 --> 00:15:23.000 system. It is. If a company's expense ratio is 00:15:23.000 --> 00:15:25.980 trending upward over a few quarters, it signals 00:15:25.980 --> 00:15:29.029 a structural problem. Either revenue is stalling 00:15:29.029 --> 00:15:31.450 or the operational costs are getting bloated 00:15:31.450 --> 00:15:34.490 relative to the scale of sales. Okay, we've thoroughly 00:15:34.490 --> 00:15:37.789 nailed down the cost side, that predictable structural 00:15:37.789 --> 00:15:40.889 drag represented by the expense ratio. We know 00:15:40.889 --> 00:15:43.389 it's sticky, disproportionately affects low -yield 00:15:43.389 --> 00:15:45.850 assets, and can be hidden through recoupment 00:15:45.850 --> 00:15:48.889 plans. Now let's pivot entirely to the reward 00:15:48.889 --> 00:15:51.889 -for -risk side and move into the Sharpe ratio. 00:15:52.429 --> 00:15:54.889 Perfect. So if the expense ratio is the concrete 00:15:54.889 --> 00:15:57.350 cost you pay for the journey, the Sharpe ratio 00:15:57.350 --> 00:15:59.370 is the metric that tells you, mathematically, 00:15:59.809 --> 00:16:02.769 how efficiently the manager used risk to generate 00:16:02.769 --> 00:16:05.250 returns. It's the measure that tries to differentiate 00:16:05.250 --> 00:16:07.970 skill from just sheer luck or from taking on 00:16:07.970 --> 00:16:09.610 way too much risk. What's its core function? 00:16:10.059 --> 00:16:11.860 Its core function is to measure the performance 00:16:11.860 --> 00:16:14.120 of an investment, a security, or a portfolio 00:16:14.120 --> 00:16:17.019 compared to a risk -free asset. But critically, 00:16:17.259 --> 00:16:19.580 it adjusts for total risk. It quantifies the 00:16:19.580 --> 00:16:21.620 premium return you receive for every unit of 00:16:21.620 --> 00:16:24.360 volatility you take on. And the measure itself 00:16:24.360 --> 00:16:26.799 is named after the Nobel laureate William F. 00:16:26.820 --> 00:16:30.419 Sharp, who developed it back in 1966. What should 00:16:30.419 --> 00:16:32.000 our listener understand about its historical 00:16:32.000 --> 00:16:34.940 context? Well... The initial version was called 00:16:34.940 --> 00:16:37.480 the reward to variability ratio, and it used 00:16:37.480 --> 00:16:40.580 a constant risk free rate as the benchmark. But 00:16:40.580 --> 00:16:42.960 Sharpe himself revised the definition in 1994, 00:16:43.340 --> 00:16:45.919 which is really key historical detail. And what 00:16:45.919 --> 00:16:48.659 made that revision necessary? The original static 00:16:48.659 --> 00:16:51.080 risk free rate. It just didn't account for changing 00:16:51.080 --> 00:16:54.100 market dynamics. If you were comparing two funds 00:16:54.100 --> 00:16:56.559 across a decade where interest rates were changing 00:16:56.559 --> 00:17:00.559 wildly, say the volatile 1970s and 80s, that 00:17:00.559 --> 00:17:03.860 static risk free rate completely invalid. Because 00:17:03.860 --> 00:17:06.700 the opportunity cost of capital was always shifting. 00:17:06.839 --> 00:17:10.420 Exactly. The 1994 revision replaced that constant 00:17:10.420 --> 00:17:12.660 risk -free rate with a more applicable and time 00:17:12.660 --> 00:17:14.900 -changing benchmark return. And that made the 00:17:14.900 --> 00:17:17.359 ratio far more robust for long -term comparative 00:17:17.359 --> 00:17:20.220 analysis. Let's look at the modern formula, even 00:17:20.220 --> 00:17:22.380 if just conceptually, to really understand the 00:17:22.380 --> 00:17:24.140 relationship between the numerator, which is 00:17:24.140 --> 00:17:26.960 the reward, and the denominator, the risk. We 00:17:26.960 --> 00:17:29.380 typically look at the ex -ante formula. The one 00:17:29.380 --> 00:17:31.700 using expected returns. But the concept holds 00:17:31.700 --> 00:17:34.819 for ex post or realized returns as well. The 00:17:34.819 --> 00:17:37.140 numerator is the expected value of the excess 00:17:37.140 --> 00:17:40.200 return. So that's the assets return minus the 00:17:40.200 --> 00:17:42.789 benchmark return. which is often the U .S. Treasury 00:17:42.789 --> 00:17:45.210 security. Correct. It's measuring the premium, 00:17:45.329 --> 00:17:47.470 the return that exceeded what you could have 00:17:47.470 --> 00:17:50.369 gotten in a virtually risk -free vehicle. And 00:17:50.369 --> 00:17:53.150 importantly, that asset return is the net return. 00:17:53.289 --> 00:17:56.170 It's calculated after the expense ratio has been 00:17:56.170 --> 00:17:58.269 deducted. And that's a crucial link back to part 00:17:58.269 --> 00:18:01.950 one. It is. If your ER is high, it drags down 00:18:01.950 --> 00:18:04.210 that asset return and automatically reduces your 00:18:04.210 --> 00:18:06.309 excess return. That reduces your Sharpe ratio, 00:18:06.470 --> 00:18:08.529 regardless of the manager's skill. So the expense 00:18:08.529 --> 00:18:11.170 ratio is guaranteed to improve. impact the Sharpe 00:18:11.170 --> 00:18:14.609 ratio negatively. Now the denominator. This is 00:18:14.609 --> 00:18:16.450 the critical risk adjustment. The denominator 00:18:16.450 --> 00:18:18.910 is the standard deviation of the asset's excess 00:18:18.910 --> 00:18:22.369 return. This is the key measure of risk, or volatility. 00:18:23.130 --> 00:18:25.609 Standard deviation measures the historical spread 00:18:25.609 --> 00:18:28.680 of returns around the average. By dividing the 00:18:28.680 --> 00:18:30.900 excess return by the standard deviation of that 00:18:30.900 --> 00:18:33.880 excess return, you quantify exactly how much 00:18:33.880 --> 00:18:36.460 bang for your risk buck you received. So a higher 00:18:36.460 --> 00:18:38.680 Sharpe ratio means you got more premium return 00:18:38.680 --> 00:18:40.400 for the amount of volatility you had to endure. 00:18:40.599 --> 00:18:42.720 That's it. When you're comparing two assets, 00:18:42.960 --> 00:18:45.799 if asset A has a Sharpe ratio of 1 .0 and asset 00:18:45.799 --> 00:18:49.740 B has 0 .5, you should prefer A because it generated 00:18:49.740 --> 00:18:52.599 twice the reward for the same unit of risk. That's 00:18:52.599 --> 00:18:55.279 the textbook answer anyway. It is, and the real 00:18:55.279 --> 00:18:57.440 -world evidence can be pretty compelling. The 00:18:57.440 --> 00:18:59.559 source provides the example of Berkshire Hathaway, 00:18:59.740 --> 00:19:02.480 which achieved a sharp ratio of 0 .79 between 00:19:02.480 --> 00:19:06.019 1976 and 2017. You compare that to the overall 00:19:06.019 --> 00:19:08.799 stock market, the S &P 500, which only managed 00:19:08.799 --> 00:19:12.099 0 .49 during that same period. So that 0 .79 00:19:12.099 --> 00:19:14.220 number isn't just about getting high returns. 00:19:14.339 --> 00:19:16.480 It means that Warren Buffett generated those 00:19:16.480 --> 00:19:19.000 high returns while incurring significantly less 00:19:19.000 --> 00:19:21.099 year -to -year volatility than the market itself. 00:19:21.680 --> 00:19:23.880 That's really the definition of a high -quality 00:19:23.880 --> 00:19:27.019 return. It demonstrates superior risk management. 00:19:27.460 --> 00:19:29.519 And before we move on, we should just briefly 00:19:29.519 --> 00:19:32.799 touch on the information ratio. While the Sharpe 00:19:32.799 --> 00:19:35.579 ratio uses the risk -free rate as the benchmark, 00:19:35.900 --> 00:19:38.500 the information ratio substitutes a typically 00:19:38.500 --> 00:19:41.619 risky index, like a specific sector index, for 00:19:41.619 --> 00:19:43.920 that benchmark. So it's for measuring a manager 00:19:43.920 --> 00:19:47.319 against their specific mandate. Exactly. Now, 00:19:47.339 --> 00:19:49.339 Sharpe ratio stands on the shoulders of earlier 00:19:49.339 --> 00:19:51.970 thought. Let's look at the precursors, specifically 00:19:51.970 --> 00:19:55.710 Andrew D. Roy's ratio from back in 1952. Roy's 00:19:55.710 --> 00:19:58.849 ratio was conceptually similar. It aimed to maximize 00:19:58.849 --> 00:20:01.230 a ratio, but the key philosophical difference 00:20:01.230 --> 00:20:04.049 was in the numerator. Roy used what he called 00:20:04.049 --> 00:20:06.470 the disaster level, or the minimum acceptable 00:20:06.470 --> 00:20:09.490 return, MAR, rather than the risk -free rate. 00:20:09.710 --> 00:20:12.369 So he was focused on avoiding a specific level 00:20:12.369 --> 00:20:14.549 of loss. Right. He was focused on the probability 00:20:14.549 --> 00:20:16.970 of a return falling below a specified acceptable 00:20:16.970 --> 00:20:19.539 threshold. And the denominator was also a bit 00:20:19.539 --> 00:20:21.539 different. He used the standard deviation of 00:20:21.539 --> 00:20:23.900 gross returns, whereas Sharpe's modern ratio 00:20:23.900 --> 00:20:26.180 uses the standard deviation of excess returns. 00:20:26.460 --> 00:20:29.099 It's a subtle but important shift in focus. And 00:20:29.099 --> 00:20:31.440 this discussion about a minimum acceptable return, 00:20:31.720 --> 00:20:34.539 it immediately brings up the Sortino ratio, which 00:20:34.539 --> 00:20:37.079 is often cited as a superior measure in some 00:20:37.079 --> 00:20:39.680 contexts. Well, the Sortino ratio is philosophically 00:20:39.680 --> 00:20:41.880 aligned with Roy's initial concept because it 00:20:41.880 --> 00:20:44.880 also uses that MAR in the numerator. But the 00:20:44.880 --> 00:20:47.420 key technical advantage of Sortino is in the 00:20:47.420 --> 00:20:50.799 denominator. It uses downside deviation. Meaning 00:20:50.799 --> 00:20:54.039 it only penalizes volatility when returns fall 00:20:54.039 --> 00:20:56.759 below that minimum acceptable return. Exactly. 00:20:56.779 --> 00:20:59.420 It ignores upside volatility, which most investors 00:20:59.420 --> 00:21:01.779 view as a good thing. Sharp, on the other hand, 00:21:01.839 --> 00:21:04.559 penalizes all volatility equally, up or down. 00:21:04.940 --> 00:21:06.779 That's a crucial distinction, and we'll come 00:21:06.779 --> 00:21:09.079 back to that in the outro. Finally, the necessary 00:21:09.079 --> 00:21:11.720 comparison, Sharpe versus the trainer ratio. 00:21:12.140 --> 00:21:14.500 The distinction is fundamental. It relates to 00:21:14.500 --> 00:21:16.420 the type of risk the manager is being measured 00:21:16.420 --> 00:21:19.700 on. The Sharpe ratio considers total risk, both 00:21:19.700 --> 00:21:22.359 systematic risk, which is market risk you can't 00:21:22.359 --> 00:21:25.160 diversify away, and idiosyncratic risk, which 00:21:25.160 --> 00:21:27.460 is asset -specific risk you can diversify away. 00:21:27.660 --> 00:21:30.059 And the trainer ratio. The trainer ratio considers 00:21:30.059 --> 00:21:33.319 only systematic risk. And this is because the 00:21:33.319 --> 00:21:35.759 trainer formula divides the excess return by 00:21:35.759 --> 00:21:39.000 beta. And beta is inherently a measure of systematic 00:21:39.000 --> 00:21:41.980 risk, the asset sensitivity to the overall market 00:21:41.980 --> 00:21:45.240 movement. So if I am building a portfolio with, 00:21:45.319 --> 00:21:48.599 say, 100 different stocks, my portfolio is already 00:21:48.599 --> 00:21:52.359 highly diversified. In that case, the idiosyncratic 00:21:52.359 --> 00:21:55.140 risk, the company -specific risk, has largely 00:21:55.140 --> 00:21:58.319 been filtered out. Exactly. So if you, the investor, 00:21:58.480 --> 00:22:01.000 are looking at a manager running an already well 00:22:01.000 --> 00:22:03.359 -diversified portfolio that is just one component 00:22:03.359 --> 00:22:05.900 of your overall assets, trainer is often the 00:22:05.900 --> 00:22:08.740 superior metric. It rewards the manager only 00:22:08.740 --> 00:22:11.420 for the market risk they must take. Sharp is 00:22:11.420 --> 00:22:14.119 generally better for evaluating a single, undiversified 00:22:14.119 --> 00:22:17.079 asset or a strategy in isolation. And since these 00:22:17.079 --> 00:22:19.920 measures rely so heavily on math, they need statistical 00:22:19.920 --> 00:22:22.339 validation. That's right. There are statistical 00:22:22.339 --> 00:22:24.960 tests like those proposed by Jobson and Corky 00:22:24.960 --> 00:22:27.319 and by Gibbons, Ross and Schenken. They exist 00:22:27.319 --> 00:22:29.500 to determine the statistical significance of 00:22:29.500 --> 00:22:32.000 a high Sharpe ratio, helping investors figure 00:22:32.000 --> 00:22:34.380 out whether a manager's success is due to genuine, 00:22:34.480 --> 00:22:37.579 repeatable skill or just statistical luck over 00:22:37.579 --> 00:22:40.779 a short, favorable period. This section is the 00:22:40.779 --> 00:22:43.940 Deep Dive's core warning label. The Sharpe ratio 00:22:43.940 --> 00:22:46.819 is a powerful tool, but it rests on a foundational 00:22:46.819 --> 00:22:49.809 assumption about the market that is. Well, it's 00:22:49.809 --> 00:22:52.230 often violated. It is. It assumes asset returns 00:22:52.230 --> 00:22:54.549 are normally distributed following that classic 00:22:54.549 --> 00:22:57.049 neat bell curve. And the reality is that the 00:22:57.049 --> 00:23:00.349 market rarely, if ever, conforms to a perfect 00:23:00.349 --> 00:23:03.180 bell curve. Precisely. Asset returns exhibit 00:23:03.180 --> 00:23:06.259 abnormalities. We see skewness where the distribution 00:23:06.259 --> 00:23:09.680 is lopsided, maybe favoring modest positive returns, 00:23:09.900 --> 00:23:12.460 but having a long, thin tail on the catastrophic 00:23:12.460 --> 00:23:15.480 loss side. And we also see kurtosis, which means 00:23:15.480 --> 00:23:17.700 the distribution has higher peaks around the 00:23:17.700 --> 00:23:20.319 average and fatter tails than predicted by a 00:23:20.319 --> 00:23:22.720 standard bell curve. Let's clarify fatter tails. 00:23:22.839 --> 00:23:25.069 What does that mean for an investor? Fatter tails 00:23:25.069 --> 00:23:28.329 mean that extreme rare events, both massive gains 00:23:28.329 --> 00:23:30.529 and huge losses, they occur more frequently than 00:23:30.529 --> 00:23:32.230 the mathematical model of the normal distribution 00:23:32.230 --> 00:23:35.150 would predict. The standard deviation, which 00:23:35.150 --> 00:23:37.329 is the denominator in the Charp ratio, it becomes 00:23:37.329 --> 00:23:39.950 an ineffective measure of risk when these abnormalities 00:23:39.950 --> 00:23:43.210 exist. It will tell you the average routine volatility 00:23:43.210 --> 00:23:45.430 is low, giving you a false sense of security. 00:23:45.960 --> 00:23:48.240 The danger is that the standard deviation calculated 00:23:48.240 --> 00:23:51.740 on all that routine data fails to capture the 00:23:51.740 --> 00:23:54.640 true magnitude and probability of those rare 00:23:54.640 --> 00:23:58.420 catastrophic losses, the black swan events. You've 00:23:58.420 --> 00:24:00.980 hit the nail on the head. If a strategy's returns 00:24:00.980 --> 00:24:04.079 are negatively skewed, meaning losses when they 00:24:04.079 --> 00:24:07.140 occur are huge and rare, the standard deviation 00:24:07.140 --> 00:24:10.160 calculated over a normal period will be deceptively 00:24:10.160 --> 00:24:13.539 low. That artificially inflates the Sharpe ratio 00:24:13.539 --> 00:24:16.640 and compromises its accuracy as a true predictor 00:24:16.640 --> 00:24:18.920 of downside risk. This inherent vulnerability 00:24:18.920 --> 00:24:21.720 means the Sharpe ratio can be manipulated, or 00:24:21.720 --> 00:24:24.599 at least strategically exploited, by fund managers 00:24:24.599 --> 00:24:27.339 who are focused on optimizing the number rather 00:24:27.339 --> 00:24:30.259 than than minimizing the actual risk. The examples 00:24:30.259 --> 00:24:32.519 here are deeply instructive. The classic example 00:24:32.519 --> 00:24:35.099 is the Ponzi scheme. Before it collapses, a Ponzi 00:24:35.099 --> 00:24:37.440 scheme generates these stable, positive reported 00:24:37.440 --> 00:24:39.839 returns because the returns are just earlier 00:24:39.839 --> 00:24:42.670 investors principle. Since the reported returns 00:24:42.670 --> 00:24:45.150 are so stable, the standard deviation is virtually 00:24:45.150 --> 00:24:47.450 zero. Which gives it an incredibly high Sharpe 00:24:47.450 --> 00:24:50.029 ratio. An unbelievably high empirical Sharpe 00:24:50.029 --> 00:24:52.190 ratio. The number looks fantastic, but it's masking 00:24:52.190 --> 00:24:55.029 a guaranteed 100 % loss risk. That shows the 00:24:55.029 --> 00:24:57.750 SR is only as good as the underlying data reflecting 00:24:57.750 --> 00:25:01.430 genuine market risk. What about legal strategies 00:25:01.430 --> 00:25:04.490 that similarly exploit this flaw? A very common 00:25:04.490 --> 00:25:07.130 one is selling low strike put options or out 00:25:07.130 --> 00:25:09.450 of the money options. This generates a small, 00:25:09.549 --> 00:25:12.049 consistent premium income, which keeps returns 00:25:12.049 --> 00:25:16.210 stable, positive and volatility low. A manager 00:25:16.210 --> 00:25:18.690 can run this strategy for years and produce a 00:25:18.690 --> 00:25:21.569 superb sharp ratio. But this strategy carries 00:25:21.569 --> 00:25:24.390 a small probability of a catastrophic, ruinous 00:25:24.390 --> 00:25:26.750 loss if the market crashes and those puts are 00:25:26.750 --> 00:25:29.829 exercised. So the SR measures the frequency and 00:25:29.829 --> 00:25:33.029 size of all the small winning moves, but it utterly 00:25:33.200 --> 00:25:35.180 fails to weigh the potential for that single 00:25:35.180 --> 00:25:37.940 ruinous outlier event that defines the strategy's 00:25:37.940 --> 00:25:40.380 actual risk. Precisely. Another vulnerability 00:25:40.380 --> 00:25:42.759 comes up with illiquid assets where the SR is 00:25:42.759 --> 00:25:44.819 susceptible to manipulation through smoothing 00:25:44.819 --> 00:25:47.319 returns. If a fund can use discretionary pricing 00:25:47.319 --> 00:25:49.619 or internal valuation models for assets that 00:25:49.619 --> 00:25:51.819 don't trade daily, they can artificially smooth 00:25:51.819 --> 00:25:53.920 out the volatility. They suppress the standard 00:25:53.920 --> 00:25:56.680 deviation and inflate the SR. So smoothing returns 00:25:56.680 --> 00:25:59.140 is essentially just taking the peaks and valleys 00:25:59.140 --> 00:26:01.990 out of the performance to make. the volatility 00:26:01.990 --> 00:26:04.710 look lower than it truly is. That's right. For 00:26:04.710 --> 00:26:06.829 assets with smoothing, like certain real estate 00:26:06.829 --> 00:26:09.109 funds or private equity vehicles, the reported 00:26:09.109 --> 00:26:12.509 fund returns are inherently misleading. The SR 00:26:12.509 --> 00:26:14.549 should ideally be derived from the performance 00:26:14.549 --> 00:26:17.250 of the underlying assets. The actual cash flow 00:26:17.250 --> 00:26:20.069 or market price changes, not just the reported 00:26:20.069 --> 00:26:22.829 fund returns. And are there statistical tools 00:26:22.829 --> 00:26:26.450 available to detect managers who might be artificially 00:26:26.450 --> 00:26:29.279 smoothing volatility? There are. Analysts look 00:26:29.279 --> 00:26:32.119 at statistics like the bias ratio and first order 00:26:32.119 --> 00:26:35.599 autocorrelation. High autocorrelation, where 00:26:35.599 --> 00:26:37.799 the return in one period is highly correlated 00:26:37.799 --> 00:26:40.460 with the return in the next, is often a statistical 00:26:40.460 --> 00:26:42.599 red flag that returns are being artificially 00:26:42.599 --> 00:26:44.960 managed or smoothed, as true market movements 00:26:44.960 --> 00:26:47.099 are rarely that perfectly connected day to day. 00:26:47.180 --> 00:26:49.160 And finally, there's the need for data length. 00:26:49.549 --> 00:26:52.930 If the ruinous events are rare, you need a commensurately 00:26:52.930 --> 00:26:55.769 long track record to capture them. A reliable 00:26:55.769 --> 00:26:58.809 empirical SR estimate requires collecting data 00:26:58.809 --> 00:27:01.450 over a sufficient period to observe all aspects 00:27:01.450 --> 00:27:04.210 of the strategy. If a high -frequency trading 00:27:04.210 --> 00:27:06.789 algorithm makes thousands of trades daily, you 00:27:06.789 --> 00:27:09.589 might get a reliable SR in a month. But if a 00:27:09.589 --> 00:27:11.589 strategy is built around catastrophic insurance 00:27:11.589 --> 00:27:13.750 liabilities that only pay out once a decade, 00:27:14.009 --> 00:27:16.630 you need decades of data to truly understand 00:27:16.630 --> 00:27:18.670 the risk. And even then you might miss something. 00:27:19.069 --> 00:27:21.950 You might. The source emphasizes that even long 00:27:21.950 --> 00:27:24.930 track records may still miss those rare extreme 00:27:24.930 --> 00:27:28.170 risks if the strategy has a very long tail of 00:27:28.170 --> 00:27:30.769 potential loss. Let's move to how the ratio behaves 00:27:30.769 --> 00:27:33.630 in scenarios where performance is just poor or 00:27:33.630 --> 00:27:36.029 the incentives are misaligned. What happens when 00:27:36.029 --> 00:27:38.269 the Sharpe ratio turns negative? A negative Sharpe 00:27:38.269 --> 00:27:40.789 ratio means the portfolio underperformed its 00:27:40.789 --> 00:27:43.140 benchmark. likely because the excess return, 00:27:43.400 --> 00:27:46.160 the numerator, is negative. And when the ratio 00:27:46.160 --> 00:27:48.380 is negative, it becomes deeply counterintuitive. 00:27:48.619 --> 00:27:50.559 Can you walk us through that specific mathematical 00:27:50.559 --> 00:27:53.500 oddity? If the SR is negative and the manager 00:27:53.500 --> 00:27:55.980 were to suddenly increase the volatility, the 00:27:55.980 --> 00:27:58.680 denominator, which is a bad thing, the denominator 00:27:58.680 --> 00:28:02.279 gets larger. Because the ratio is negative, dividing 00:28:02.279 --> 00:28:04.839 by a larger positive number actually makes the 00:28:04.839 --> 00:28:07.460 resulting negative ratio numerically less negative. 00:28:07.579 --> 00:28:09.319 And therefore technically higher. Technically 00:28:09.319 --> 00:28:11.779 higher. That is deeply confusing. So increasing 00:28:11.779 --> 00:28:14.140 risk makes a bad number look technically better. 00:28:14.380 --> 00:28:16.960 If I see a manager reporting an SR of native 00:28:16.960 --> 00:28:21.140 0 .5 and another reporting native 0 .1, I'm supposed 00:28:21.140 --> 00:28:24.019 to prefer the native 0 .1 even if the only way 00:28:24.019 --> 00:28:26.299 they got there was by increasing their risk. 00:28:26.700 --> 00:28:29.440 This proves the SR is often useless in bear markets 00:28:29.440 --> 00:28:31.720 or for strategies that are underperforming the 00:28:31.720 --> 00:28:34.720 risk -free rate. Investor utility functions always 00:28:34.720 --> 00:28:37.200 prefer lower volatility, but the math of the 00:28:37.200 --> 00:28:39.880 negative SR rewards higher volatility. That's 00:28:39.880 --> 00:28:42.480 why many analysts only use the SR when the excess 00:28:42.480 --> 00:28:45.500 return is positive. This fragility also contributes 00:28:45.500 --> 00:28:47.980 directly to the principal -agent problem. The 00:28:47.980 --> 00:28:50.700 investor, the principal, wants safety and sustainable 00:28:50.700 --> 00:28:53.680 returns, but the manager, the agent, might be 00:28:53.680 --> 00:28:56.109 incentivized to just chase a high number. That's 00:28:56.109 --> 00:28:58.910 a textbook misalignment. Fund sponsors who select 00:28:58.910 --> 00:29:01.190 managers solely on who generates the highest 00:29:01.190 --> 00:29:04.410 SR can unintentionally incentivize those managers 00:29:04.410 --> 00:29:08.069 to adopt high SR, high risk strategies, like 00:29:08.069 --> 00:29:10.950 that short option strategy we mentioned. Selling 00:29:10.950 --> 00:29:13.230 options generates consistent fees and modest 00:29:13.230 --> 00:29:15.950 positive payoffs, which leads to a great SR, 00:29:16.150 --> 00:29:18.589 which maximizes the manager's short term bonus. 00:29:18.809 --> 00:29:21.509 So the manager gets paid handsomely. based on 00:29:21.509 --> 00:29:24.369 the high SR, but the investor holds the bag for 00:29:24.369 --> 00:29:26.430 that small probability of catastrophic loss. 00:29:27.000 --> 00:29:29.700 Exactly. The source specifically notes that selling 00:29:29.700 --> 00:29:31.920 out of the money calls and puts is mathematically 00:29:31.920 --> 00:29:34.700 determined to maximize the SR in the short term, 00:29:34.799 --> 00:29:37.440 even though it exposes the fund to ruinous tail 00:29:37.440 --> 00:29:39.359 risk that most investors are trying to avoid. 00:29:39.599 --> 00:29:41.940 This perverse incentive structure is a fundamental 00:29:41.940 --> 00:29:44.339 failing of relying solely on the standard SR 00:29:44.339 --> 00:29:46.599 for manager selection. And finally, we have to 00:29:46.599 --> 00:29:49.240 address that common oversimplified rubric we 00:29:49.240 --> 00:29:51.200 see in financial media. The one that says an 00:29:51.200 --> 00:29:53.619 SR over one is acceptable, over two is very good, 00:29:53.660 --> 00:29:56.089 and over three is excellent. Why is this rubric 00:29:56.089 --> 00:29:58.890 fundamentally flawed? It is flawed because a 00:29:58.890 --> 00:30:02.109 Sharpe ratio is a dimensional quantity. Its magnitude 00:30:02.109 --> 00:30:04.730 is highly sensitive to the time period over which 00:30:04.730 --> 00:30:07.170 the returns are measured. The comparison only 00:30:07.170 --> 00:30:10.069 works if the time frame is standardized, typically 00:30:10.069 --> 00:30:12.900 annualized. explain the math behind that scaling 00:30:12.900 --> 00:30:15.859 problem the numerator the expected excess return 00:30:15.859 --> 00:30:19.480 it scales linearly with time if you double the 00:30:19.480 --> 00:30:22.059 time period you double the expected return however 00:30:22.059 --> 00:30:25.640 the denominator the standard deviation scales 00:30:25.640 --> 00:30:28.019 with the square root of time So if I calculate 00:30:28.019 --> 00:30:31.519 a monthly Sharpe ratio of 0 .3, and I incorrectly 00:30:31.519 --> 00:30:34.000 assume I can just multiply the entire ratio by 00:30:34.000 --> 00:30:35.980 12 to annualize it, I'm going to get a really 00:30:35.980 --> 00:30:38.140 inflated number. You are. The return portion 00:30:38.140 --> 00:30:40.720 scales by 12, but the risk portion only scales 00:30:40.720 --> 00:30:43.380 by the square root of 12, which is about 3 .46. 00:30:44.000 --> 00:30:46.640 You're understating the annualized risk relative 00:30:46.640 --> 00:30:49.099 to the annualized return, and that inflates the 00:30:49.099 --> 00:30:51.720 final ratio. The accepted method requires you 00:30:51.720 --> 00:30:53.660 to first annualize the return and the standard 00:30:53.660 --> 00:30:55.799 deviation separately, using the right scaling 00:30:55.799 --> 00:30:58.920 factors. and then calculate the ratio. So those 00:30:58.920 --> 00:31:01.480 targets of 2 .0 or 3 .0 are pretty unrealistic. 00:31:01.900 --> 00:31:04.380 For most diversified asset classes, they're utterly 00:31:04.380 --> 00:31:07.819 unrealistic. Most diversified indexes have annualized 00:31:07.819 --> 00:31:10.720 sharp ratios below one. Given all these inherent 00:31:10.720 --> 00:31:13.119 weaknesses, analysts have developed several advanced 00:31:13.119 --> 00:31:16.079 methods either to correct the SR or provide alternatives. 00:31:16.680 --> 00:31:19.119 What are the major corrections being used today? 00:31:19.450 --> 00:31:22.069 For practitioners, especially those dealing with 00:31:22.069 --> 00:31:24.309 shorter track records and potential data issues, 00:31:24.609 --> 00:31:28.269 the deflated Sharpe ratio is critical. This ratio 00:31:28.269 --> 00:31:30.829 was proposed specifically for hedge funds with 00:31:30.829 --> 00:31:33.670 short track records, and it aims to correct for 00:31:33.670 --> 00:31:36.509 asymmetry, fat tails, sample length, and the 00:31:36.509 --> 00:31:38.809 selection bias that tends to make reported SRs 00:31:38.809 --> 00:31:41.630 look overly optimistic. So it's a stress -tested 00:31:41.630 --> 00:31:43.410 Sharpe ratio. That's a good way to put it. It's 00:31:43.410 --> 00:31:46.049 an SR that has been rigorously stress -tested 00:31:46.049 --> 00:31:48.589 for all these known flaws. What about the application? 00:31:48.779 --> 00:31:51.819 side, how can the SR be converted into an actionable 00:31:51.819 --> 00:31:54.460 metric for sizing investments? That's where the 00:31:54.460 --> 00:31:57.759 Kelly criterion comes in. While the SR is a risk 00:31:57.759 --> 00:32:00.640 -adjusted ratio, the Kelly criterion can be used 00:32:00.640 --> 00:32:03.019 to convert it into a suggested rate of return 00:32:03.019 --> 00:32:06.339 based on optimal sizing. It calculates the ideal 00:32:06.339 --> 00:32:08.759 size of an investment, which, when you adjust 00:32:08.759 --> 00:32:10.799 it by the period and expected rate of return 00:32:10.799 --> 00:32:13.180 per unit, gives you a prescriptive rate of return. 00:32:13.339 --> 00:32:15.660 It moves the metric from just comparative performance 00:32:15.660 --> 00:32:19.119 to optimal capital allocation. And finally, let's 00:32:19.119 --> 00:32:21.740 revisit portfolio construction. We established 00:32:21.740 --> 00:32:24.299 that idiosyncratic risk can be diversified away. 00:32:24.599 --> 00:32:26.720 Is there a scenario where you would intentionally 00:32:26.720 --> 00:32:29.759 hire a manager with a low or even a negative 00:32:29.759 --> 00:32:32.180 Sharpe ratio? Absolutely, and this is illustrated 00:32:32.180 --> 00:32:34.400 by the proposed Sharpe ratio indifference curve. 00:32:34.990 --> 00:32:37.109 This concept demonstrates that it can be highly 00:32:37.109 --> 00:32:40.329 efficient to hire managers with low or even negative 00:32:40.329 --> 00:32:43.190 individual Sharpey ratios, provided their correlation 00:32:43.190 --> 00:32:45.890 to your overall portfolio is sufficiently low. 00:32:46.089 --> 00:32:48.069 Because they zigzag when everyone else is moving 00:32:48.069 --> 00:32:50.769 in the same direction. Yes. Their poor individual 00:32:50.769 --> 00:32:54.009 SR is offset by the tremendous diversification 00:32:54.009 --> 00:32:56.009 benefit they provide to the whole portfolio. 00:32:56.410 --> 00:32:58.910 They act as an insurance policy, potentially 00:32:58.910 --> 00:33:01.390 benefiting the overall portfolio Sharpe ratio, 00:33:01.670 --> 00:33:03.670 even if their individual number looks terrible. 00:33:04.109 --> 00:33:06.390 It illustrates that good portfolio construction 00:33:06.390 --> 00:33:09.210 requires looking beyond the single SR number 00:33:09.210 --> 00:33:12.269 of any one component. So what does this all mean 00:33:12.269 --> 00:33:14.329 for you, the learner? We started with the expense 00:33:14.329 --> 00:33:17.230 ratio, that predictable, guaranteed cost that 00:33:17.230 --> 00:33:19.650 quietly consumes your net returns, sometimes 00:33:19.650 --> 00:33:22.130 disproportionately, especially in those low -yield 00:33:22.130 --> 00:33:24.950 vehicles. And we highlighted the need to watch 00:33:24.950 --> 00:33:27.190 out for recoupment plans that hide future ER 00:33:27.190 --> 00:33:29.329 increases. Then we tackled the Sharpe ratio. 00:33:29.900 --> 00:33:32.140 a powerful but fundamentally tricky metric that's 00:33:32.140 --> 00:33:34.099 trying to quantify the quality of the return 00:33:34.099 --> 00:33:36.900 relative to the total risk taken. We uncovered 00:33:36.900 --> 00:33:39.839 that a high SR doesn't guarantee safety, as it 00:33:39.839 --> 00:33:41.839 can be the result of ignoring fat -tail risks, 00:33:42.160 --> 00:33:44.680 smoothing returns, or adopting strategies that 00:33:44.680 --> 00:33:46.619 are optimized for high short -term returns but 00:33:46.619 --> 00:33:49.140 carry catastrophic long -term risks. And what's 00:33:49.140 --> 00:33:50.839 fascinating here is the interplay between them. 00:33:51.519 --> 00:33:54.640 The expense ratio, which is concrete and predictable, 00:33:54.859 --> 00:33:57.799 directly impacts the asset return in the Sharpe 00:33:57.799 --> 00:34:01.000 ratio numerator. It is the structural cost. And 00:34:01.000 --> 00:34:03.519 if your costs are excessively high, no matter 00:34:03.519 --> 00:34:05.519 how clever your manager is at risk selection, 00:34:05.859 --> 00:34:08.860 that cost will suppress your net return, reducing 00:34:08.860 --> 00:34:11.320 your excess return and, well, resulting in a 00:34:11.320 --> 00:34:14.179 lower Sharpe ratio. They are inextricably linked. 00:34:14.400 --> 00:34:17.780 The ER determines the certain cost, and SR measures 00:34:17.780 --> 00:34:20.320 the efficiency of the return generated after 00:34:20.320 --> 00:34:22.659 that cost is paid. The key is knowing where the 00:34:22.659 --> 00:34:25.170 metrics are vulnerable. Never trust the number 00:34:25.170 --> 00:34:27.110 without understanding the time period used for 00:34:27.110 --> 00:34:29.010 the calculation and the assumptions about return 00:34:29.010 --> 00:34:31.690 normality. For managers who are paid based on 00:34:31.690 --> 00:34:34.469 optimizing a number, you must always verify the 00:34:34.469 --> 00:34:36.409 type of risk they are actually running. Which 00:34:36.409 --> 00:34:39.110 raises an important final question. Given the 00:34:39.110 --> 00:34:41.389 known vulnerabilities of the Sharpe ratio to 00:34:41.389 --> 00:34:43.849 non -normal distributions and its failure to 00:34:43.849 --> 00:34:46.190 distinguish between upside and downside volatility, 00:34:46.530 --> 00:34:48.949 how might a reliance on metrics that specifically 00:34:48.949 --> 00:34:52.030 penalize only downside risk, like the Sortino 00:34:52.030 --> 00:34:54.590 ratio or the Omega ratio, fundamentally change 00:34:54.590 --> 00:34:56.550 the investment strategies managers choose to 00:34:56.550 --> 00:34:58.929 implement and the types of assets they buy? That's 00:34:58.929 --> 00:35:00.889 something to mull over as you analyze your next 00:35:00.889 --> 00:35:01.650 financial statement.